archive/blender-archive
Archived
1
1
Fork 0
Code Pull Requests Packages Activity
This repository has been archived on 2023-10-09. You can view files and clone it, but cannot push or open issues or pull requests.
Files
c40cc0bf6d93bde49f1ae16a663cb504d1a588d3
blender-archive/source/blender/render/intern/source/renderdatabase.c

1496 lines
37 KiB
C
Raw Normal View History

doxygen: prevent GPL license block from being parsed as doxygen comment.
2011-02-23 10:52:22 +00:00
/*
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
* ***** BEGIN GPL LICENSE BLOCK *****
Initial revision
2002-10-12 11:37:38 +00:00
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
* of the License, or (at your option) any later version.
Initial revision
2002-10-12 11:37:38 +00:00
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software Foundation,
correct fsf address
2010-02-12 13:34:04 +00:00
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
Initial revision
2002-10-12 11:37:38 +00:00
*
* The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
* All rights reserved.
*
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
* Contributor(s): 2004-2006, Blender Foundation, full recode
Initial revision
2002-10-12 11:37:38 +00:00
*
* ***** END GPL/BL DUAL LICENSE BLOCK *****
*/
doxygen: blender/render tagged.
2011-02-27 19:31:27 +00:00
/** \file blender/render/intern/source/renderdatabase.c
* \ingroup render
*/
Initial revision
2002-10-12 11:37:38 +00:00
/*
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
* Storage, retrieval and query of render specific data.
*
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
* All data from a Blender scene is converted by the renderconverter/
Initial revision
2002-10-12 11:37:38 +00:00
* into a special format that is used by the render module to make
* images out of. These functions interface to the render-specific
* database.
*
* The blo{ha/ve/vl} arrays store pointers to blocks of 256 data
* entries each.
*
* The index of an entry is >>8 (the highest 24 * bits), to find an
* offset in a 256-entry block.
*
* - If the 256-entry block entry has an entry in the
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
* vertnodes/vlaknodes/bloha array of the current block, the i-th entry in
Initial revision
2002-10-12 11:37:38 +00:00
* that block is allocated to this entry.
*
* - If the entry has no block allocated for it yet, memory is
* allocated.
*
code cleanup: spelling
2012-07-16 23:23:33 +00:00
* The pointer to the correct entry is returned. Memory is guaranteed
Initial revision
2002-10-12 11:37:38 +00:00
* to exist (as long as the malloc does not break). Since guarded
* allocation is used, memory _must_ be available. Otherwise, an
* exit(0) would occur.
*
*/
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
#include <limits.h>
Initial revision
2002-10-12 11:37:38 +00:00
#include <math.h>
Phew, a lot of work, and no new features... Main target was to make the inner rendering loop using no globals anymore. This is essential for proper usage while raytracing, it caused a lot of hacks in the raycode as well, which even didn't work correctly for all situations (textures especially). Done this by creating a new local struct RenderInput, which replaces usage of the global struct Render R. The latter now only is used to denote image size, viewmatrix, and the like. Making the inner render loops using no globals caused 1000s of vars to be changed... but the result definitely is much nicer code, which enables making 'real' shaders in a next stage. It also enabled me to remove the hacks from ray.c Then i went to the task of removing redundant code. Especially the calculus of texture coords took place (identical) in three locations. Most obvious is the change in the unified render part, which is much less code now; it uses the same rendering routines as normal render now. (Note; not for halos yet!) I also removed 6 files called 'shadowbuffer' something. This was experimen- tal stuff from NaN days. And again saved a lot of double used code. Finally I went over the blenkernel and blender/src calls to render stuff. Here the same local data is used now, resulting in less dependency. I also moved render-texture to the render module, this was still in Kernel. (new file: texture.c) So! After this commit I will check on the autofiles, to try to fix that. MSVC people have to do it themselves. This commit will need quite some testing help, but I'm around!
2003-12-21 21:52:51 +00:00
#include <string.h>
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
Initial revision
2002-10-12 11:37:38 +00:00
#include "MEM_guardedalloc.h"
remove references to BKE_utildefines where its not needed. - move GS() define into DNA_ID.h - add BLI_utildefines as an automatic include with makesrna generated files.
2011-01-07 19:18:31 +00:00
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
Math Lib * Convert all code to use new functions. * Branch maintainers may want to skip this commit, and run this conversion script instead, if they use a lot of math functions in new code: http://www.pasteall.org/9052/python
2009-11-10 20:43:45 +00:00
#include "BLI_math.h"
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
#include "BLI_blenlib.h"
split BKE_utildefines.h, now it only has blender specific defines like GS() MAKE_ID, FILE_MAXDIR, moved the generic defines to BLI_utildefines.h. no functional changes.
2011-01-07 18:36:47 +00:00
#include "BLI_utildefines.h"
Initial revision
2002-10-12 11:37:38 +00:00
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
#include "DNA_material_types.h"
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
#include "DNA_meshdata_types.h"
Initial revision
2002-10-12 11:37:38 +00:00
#include "DNA_texture_types.h"
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
#include "BKE_customdata.h"
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
#include "BKE_DerivedMesh.h"
Initial revision
2002-10-12 11:37:38 +00:00
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
#include "RE_render_ext.h" /* externtex */
updated .c files to include: #ifdef HAVE_CONFIG_H #include <config.h> #endif Just need to finish cpp files now :) Kent -- mein@cs.umn.edu
2002-11-25 12:02:15 +00:00
Fix T37675: blender internal viewport render not updating properly with volumes.
2013-12-18 15:41:28 +01:00
#include "rayintersection.h"
Raytrace modifications from the Render Branch. These should not have any effect on render results, except in some cases with you have overlapping faces, where the noise seems to be slightly reduced. There are some performance improvements, for simple scenes I wouldn't expect more than 5-10% to be cut off the render time, for sintel scenes we got about 50% on average, that's with millions of polygons on intel quad cores. This because memory access / cache misses were the main bottleneck for those scenes, and the optimizations improve that. Interal changes: * Remove RE_raytrace.h, raytracer is now only used by render engine again. * Split non-public parts rayobject.h into rayobject_internal.h, hopefully makes it clearer how the API is used. * Added rayintersection.h to contain some of the stuff from RE_raytrace.h * Change Isect.vec/labda to Isect.dir/dist, previously vec was sometimes normalized and sometimes not, confusing... now dir is always normalized and dist contains the distance. * Change VECCOPY and similar to BLI_math functions. * Force inlining of auxiliary functions for ray-triangle/quad intersection, helps a few percentages. * Reorganize svbvh code so all the traversal functions are in one file * Don't do test for root so that push_childs can be inlined * Make shadow a template parameter so it doesn't need to be runtime checked * Optimization in raytree building, was computing bounding boxes more often than necessary. * Leave out logf() factor in SAH, makes tree build quicker with no noticeable influence on raytracing on performance? * Set max childs to 4, simplifies traversal code a bit, but also seems to help slightly in general. * Store child pointers and child bb just as fixed arrays of size 4 in nodes, nearly all nodes have this many children, so overall it actually reduces memory usage a bit and avoids a pointer indirection.
2011-02-05 13:41:29 +00:00
#include "rayobject.h"
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
#include "renderpipeline.h"
#include "render_types.h"
#include "renderdatabase.h"
#include "texture.h"
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
#include "strand.h"
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
#include "zbuf.h"
Initial revision
2002-10-12 11:37:38 +00:00
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
/* ------------------------------------------------------------------------- */
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
spelling cleanup
2012-03-18 07:38:51 +00:00
/* More dynamic allocation of options for render vertices and faces, so we don't
style cleanup: comment blocks
2012-03-09 18:28:30 +00:00
* have to reserve this space inside vertices.
* Important; vertices and faces, should have been created already (to get tables
* checked) that's a reason why the calls demand VertRen/VlakRen * as arg, not
* the index */
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
/* NOTE! the hardcoded table size 256 is used still in code for going quickly over vertices/faces */
#define RE_STRESS_ELEMS 1
#define RE_RAD_ELEMS 4
#define RE_STRAND_ELEMS 1
#define RE_TANGENT_ELEMS 3
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
#define RE_WINSPEED_ELEMS 4
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
#define RE_MTFACE_ELEMS 1
#define RE_MCOL_ELEMS 4
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
#define RE_UV_ELEMS 2
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
#define RE_VLAK_ORIGINDEX_ELEMS 1
#define RE_VERT_ORIGINDEX_ELEMS 1
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
#define RE_SURFNOR_ELEMS 3
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
#define RE_RADFACE_ELEMS 1
Strand Render Simplification ============================ - Strand render now has options to remove child strands as the object's faces becomes smaller, in the Simplification particle panel. - "Reference Size" is the approximate size of the object on screen, after which simplification starts. - "Rate" is how fast strands are removed. - "Transition" is the percentage of strands being faded out as they are removed. - Another "Viewport" option removes strands on faces that are outside of the viewport. "Rate" again controls how fast these are removed. - Strand render in Blender Units now has an adjustable minimum width. Below this minimum width, strands start fading out instead of getting smaller.
2007-12-20 16:35:27 +00:00
#define RE_SIMPLIFY_ELEMS 2
Normal map tangents are now not always averaged at vertices anymore, but only when the UV's are connected. That fixes some artifacts when baking and using tangent space normal maps. It does mean increased memory usage because it now stores 4 tangents per face like UV's, and increased processing time, but there's no simple way around that.
2008-03-07 15:38:56 +00:00
#define RE_FACE_ELEMS 1
This commit will switch blender to use tangent space generated within the two files mikktspace.h and mikktspace.c. These are standalone files which can be redistributed into any other application and regenerate the same tangent spaces. The implementation is independent of the ordering of faces and the vertex ordering of faces.
2011-02-14 18:18:46 +00:00
#define RE_NMAP_TANGENT_ELEMS 16
Added Copy/Paste 'mapping' options for World/Lamp buttons. Also: moved render defines from scene to render module itself.
2005-12-03 12:55:35 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
float *RE_vertren_get_stress(ObjectRen *obr, VertRen *ver, int verify)
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
{
float *stress;
int nr= ver->index>>8;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
stress= obr->vertnodes[nr].stress;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (stress==NULL) {
if (verify)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
stress= obr->vertnodes[nr].stress= MEM_mallocN(256*RE_STRESS_ELEMS*sizeof(float), "stress table");
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
else
return NULL;
}
return stress + (ver->index & 255)*RE_STRESS_ELEMS;
}
/* this one callocs! */
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
float *RE_vertren_get_rad(ObjectRen *obr, VertRen *ver, int verify)
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
{
float *rad;
int nr= ver->index>>8;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
rad= obr->vertnodes[nr].rad;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (rad==NULL) {
if (verify)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
rad= obr->vertnodes[nr].rad= MEM_callocN(256*RE_RAD_ELEMS*sizeof(float), "rad table");
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
else
return NULL;
}
return rad + (ver->index & 255)*RE_RAD_ELEMS;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
float *RE_vertren_get_strand(ObjectRen *obr, VertRen *ver, int verify)
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
{
float *strand;
int nr= ver->index>>8;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
strand= obr->vertnodes[nr].strand;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strand==NULL) {
if (verify)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
strand= obr->vertnodes[nr].strand= MEM_mallocN(256*RE_STRAND_ELEMS*sizeof(float), "strand table");
Added Copy/Paste 'mapping' options for World/Lamp buttons. Also: moved render defines from scene to render module itself.
2005-12-03 12:55:35 +00:00
else
return NULL;
}
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
return strand + (ver->index & 255)*RE_STRAND_ELEMS;
Added Copy/Paste 'mapping' options for World/Lamp buttons. Also: moved render defines from scene to render module itself.
2005-12-03 12:55:35 +00:00
}
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
Orange; WIP commit for inclusion of "Tangent" vector in rendering. This first implementation only supports it for Meshes with UV maps, and only tangents in V direction. Tangent diffuse: http://www.blender.org/bf/0001_0080.avi Tangent spec, diffuse, and bump: http://www.blender.org/bf/20001_0080.avi NOTE: since UV coordinates are still very badly subsurfed, this won't work well for subsurf meshes... on the todo. On the todo; - generate tangents for meshes without UV (with some options) - use tangents from Curve/Surface - add the Ashkimin shader from tracker ----- Important bugfix; curves didn't render anymore since yesterday. :)
2005-12-09 15:14:32 +00:00
/* needs calloc */
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
float *RE_vertren_get_tangent(ObjectRen *obr, VertRen *ver, int verify)
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
{
float *tangent;
int nr= ver->index>>8;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
tangent= obr->vertnodes[nr].tangent;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (tangent==NULL) {
if (verify)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
tangent= obr->vertnodes[nr].tangent= MEM_callocN(256*RE_TANGENT_ELEMS*sizeof(float), "tangent table");
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
else
return NULL;
}
return tangent + (ver->index & 255)*RE_TANGENT_ELEMS;
}
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
/* needs calloc! not all renderverts have them */
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
/* also winspeed is exception, it is stored per instance */
float *RE_vertren_get_winspeed(ObjectInstanceRen *obi, VertRen *ver, int verify)
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
{
float *winspeed;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
int totvector;
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
winspeed= obi->vectors;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (winspeed==NULL) {
if (verify) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
totvector= obi->obr->totvert + obi->obr->totstrand;
winspeed= obi->vectors= MEM_callocN(totvector*RE_WINSPEED_ELEMS*sizeof(float), "winspeed table");
}
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
else
return NULL;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
return winspeed + ver->index*RE_WINSPEED_ELEMS;
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
}
Added Copy/Paste 'mapping' options for World/Lamp buttons. Also: moved render defines from scene to render module itself.
2005-12-03 12:55:35 +00:00
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
int *RE_vertren_get_origindex(ObjectRen *obr, VertRen *ver, int verify)
{
int *origindex;
int nr= ver->index>>8;
origindex= obr->vertnodes[nr].origindex;
if (origindex==NULL) {
if (verify)
origindex= obr->vertnodes[nr].origindex= MEM_mallocN(256*RE_VERT_ORIGINDEX_ELEMS*sizeof(int), "origindex table");
else
return NULL;
}
return origindex + (ver->index & 255)*RE_VERT_ORIGINDEX_ELEMS;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
VertRen *RE_vertren_copy(ObjectRen *obr, VertRen *ver)
Bugfix #5107 Autosmooth didn't copy additional texture coordinates such as sticky correctly. This was part of the vertex struct in 2.41 and older, but it became dynamic in 2.42.
2006-11-12 11:41:43 +00:00
{
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
VertRen *v1= RE_findOrAddVert(obr, obr->totvert++);
Bugfix #5107 Autosmooth didn't copy additional texture coordinates such as sticky correctly. This was part of the vertex struct in 2.41 and older, but it became dynamic in 2.42.
2006-11-12 11:41:43 +00:00
float *fp1, *fp2;
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
int *int1, *int2;
Bugfix #5107 Autosmooth didn't copy additional texture coordinates such as sticky correctly. This was part of the vertex struct in 2.41 and older, but it became dynamic in 2.42.
2006-11-12 11:41:43 +00:00
int index= v1->index;
*v1= *ver;
v1->index= index;
remove sticky coords from blender and the internal render engine.
2012-09-21 11:37:51 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
fp1= RE_vertren_get_stress(obr, ver, 0);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (fp1) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
fp2= RE_vertren_get_stress(obr, v1, 1);
Bugfix #5107 Autosmooth didn't copy additional texture coordinates such as sticky correctly. This was part of the vertex struct in 2.41 and older, but it became dynamic in 2.42.
2006-11-12 11:41:43 +00:00
memcpy(fp2, fp1, RE_STRESS_ELEMS*sizeof(float));
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
fp1= RE_vertren_get_rad(obr, ver, 0);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (fp1) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
fp2= RE_vertren_get_rad(obr, v1, 1);
Bugfix #5107 Autosmooth didn't copy additional texture coordinates such as sticky correctly. This was part of the vertex struct in 2.41 and older, but it became dynamic in 2.42.
2006-11-12 11:41:43 +00:00
memcpy(fp2, fp1, RE_RAD_ELEMS*sizeof(float));
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
fp1= RE_vertren_get_strand(obr, ver, 0);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (fp1) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
fp2= RE_vertren_get_strand(obr, v1, 1);
Bugfix #5107 Autosmooth didn't copy additional texture coordinates such as sticky correctly. This was part of the vertex struct in 2.41 and older, but it became dynamic in 2.42.
2006-11-12 11:41:43 +00:00
memcpy(fp2, fp1, RE_STRAND_ELEMS*sizeof(float));
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
fp1= RE_vertren_get_tangent(obr, ver, 0);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (fp1) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
fp2= RE_vertren_get_tangent(obr, v1, 1);
Bugfix #5107 Autosmooth didn't copy additional texture coordinates such as sticky correctly. This was part of the vertex struct in 2.41 and older, but it became dynamic in 2.42.
2006-11-12 11:41:43 +00:00
memcpy(fp2, fp1, RE_TANGENT_ELEMS*sizeof(float));
}
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
int1= RE_vertren_get_origindex(obr, ver, 0);
if (int1) {
int2= RE_vertren_get_origindex(obr, v1, 1);
memcpy(int2, int1, RE_VERT_ORIGINDEX_ELEMS*sizeof(int));
}
Bugfix #5107 Autosmooth didn't copy additional texture coordinates such as sticky correctly. This was part of the vertex struct in 2.41 and older, but it became dynamic in 2.42.
2006-11-12 11:41:43 +00:00
return v1;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
VertRen *RE_findOrAddVert(ObjectRen *obr, int nr)
Initial revision
2002-10-12 11:37:38 +00:00
{
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
VertTableNode *temp;
VertRen *v;
Initial revision
2002-10-12 11:37:38 +00:00
int a;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (nr<0) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
printf("error in findOrAddVert: %d\n", nr);
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
return NULL;
Initial revision
2002-10-12 11:37:38 +00:00
}
a= nr>>8;
Dynamic Face/Vert/Halo/Lamp tables: User Info: Hard coded limits on the total number of face, verts, halos, and lamps is gone. Blender now allocates the tables for these on an as needed basis. As long as your system can come up with the memory, you won't run out. As a bonus, it also uses slightly less memory on smaller scenes. Coder info: This has been in tuhopuu for a while, but I don't know how hard it has been tested. Since it now allocates only an initial 1024 tables (of 256 verts/faces/halos each), it seems like it has been put through it's paces. Lamps are allocated one at a time, and I start with 256. I rendered 2.5M Faces/Verts/Halos. 4444 lamps. None the less, I left a few printf's in the realocation to hunt bugs. I'll take them out just before the release freeze. Also, be on the lookout for other "sanity checks" that assume a limited number of the above items. I think I got them all, but you never know.
2003-10-19 21:47:03 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
if (a>=obr->vertnodeslen-1) { /* Need to allocate more columns..., and keep last element NULL for free loop */
temp= obr->vertnodes;
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
obr->vertnodes= MEM_mallocN(sizeof(VertTableNode)*(obr->vertnodeslen+TABLEINITSIZE), "vertnodes");
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (temp) memcpy(obr->vertnodes, temp, obr->vertnodeslen*sizeof(VertTableNode));
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
memset(obr->vertnodes+obr->vertnodeslen, 0, TABLEINITSIZE*sizeof(VertTableNode));
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->vertnodeslen+=TABLEINITSIZE;
style cleanup: trailing tabs & expand some non prefix tabs into spaces.
2012-10-21 05:46:41 +00:00
if (temp) MEM_freeN(temp);
Dynamic Face/Vert/Halo/Lamp tables: User Info: Hard coded limits on the total number of face, verts, halos, and lamps is gone. Blender now allocates the tables for these on an as needed basis. As long as your system can come up with the memory, you won't run out. As a bonus, it also uses slightly less memory on smaller scenes. Coder info: This has been in tuhopuu for a while, but I don't know how hard it has been tested. Since it now allocates only an initial 1024 tables (of 256 verts/faces/halos each), it seems like it has been put through it's paces. Lamps are allocated one at a time, and I start with 256. I rendered 2.5M Faces/Verts/Halos. 4444 lamps. None the less, I left a few printf's in the realocation to hunt bugs. I'll take them out just before the release freeze. Also, be on the lookout for other "sanity checks" that assume a limited number of the above items. I think I got them all, but you never know.
2003-10-19 21:47:03 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
v= obr->vertnodes[a].vert;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (v==NULL) {
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
int i;
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
v= (VertRen *)MEM_callocN(256*sizeof(VertRen), "findOrAddVert");
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->vertnodes[a].vert= v;
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (i= (nr & 0xFFFFFF00), a=0; a<256; a++, i++) {
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
v[a].index= i;
}
Initial revision
2002-10-12 11:37:38 +00:00
}
v+= (nr & 255);
return v;
}
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
/* ------------------------------------------------------------------------ */
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MTFace *RE_vlakren_get_tface(ObjectRen *obr, VlakRen *vlr, int n, char **name, int verify)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
{
VlakTableNode *node;
int nr= vlr->index>>8, vlakindex= (vlr->index&255);
int index= (n<<8) + vlakindex;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
node= &obr->vlaknodes[nr];
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (verify) {
if (n>=node->totmtface) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MTFace *mtface= node->mtface;
remove misc unused vars and correct theme name for face angles.
2011-01-11 07:40:31 +00:00
int size= (n+1)*256;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
node->mtface= MEM_callocN(size*sizeof(MTFace), "Vlak mtface");
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtface) {
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
size= node->totmtface*256;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
memcpy(node->mtface, mtface, size*sizeof(MTFace));
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
MEM_freeN(mtface);
}
node->totmtface= n+1;
}
}
else {
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (n>=node->totmtface)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
return NULL;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (name) *name= obr->mtface[n];
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
return node->mtface + index;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MCol *RE_vlakren_get_mcol(ObjectRen *obr, VlakRen *vlr, int n, char **name, int verify)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
{
VlakTableNode *node;
int nr= vlr->index>>8, vlakindex= (vlr->index&255);
int index= (n<<8) + vlakindex;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
node= &obr->vlaknodes[nr];
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (verify) {
if (n>=node->totmcol) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MCol *mcol= node->mcol;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
int size= (n+1)*256;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
node->mcol= MEM_callocN(size*sizeof(MCol)*RE_MCOL_ELEMS, "Vlak mcol");
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mcol) {
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
size= node->totmcol*256;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
memcpy(node->mcol, mcol, size*sizeof(MCol)*RE_MCOL_ELEMS);
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
MEM_freeN(mcol);
}
node->totmcol= n+1;
}
}
else {
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (n>=node->totmcol)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
return NULL;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (name) *name= obr->mcol[n];
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
return node->mcol + index*RE_MCOL_ELEMS;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
int *RE_vlakren_get_origindex(ObjectRen *obr, VlakRen *vlak, int verify)
{
int *origindex;
int nr= vlak->index>>8;
origindex= obr->vlaknodes[nr].origindex;
style cleanup: also correct doc example for 'foreach_get/set'
2013-02-02 00:34:34 +00:00
if (origindex==NULL) {
if (verify)
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
origindex= obr->vlaknodes[nr].origindex= MEM_callocN(256*RE_VLAK_ORIGINDEX_ELEMS*sizeof(int), "origindex table");
else
return NULL;
}
return origindex + (vlak->index & 255)*RE_VLAK_ORIGINDEX_ELEMS;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
float *RE_vlakren_get_surfnor(ObjectRen *obr, VlakRen *vlak, int verify)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
{
float *surfnor;
int nr= vlak->index>>8;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
surfnor= obr->vlaknodes[nr].surfnor;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (surfnor==NULL) {
if (verify)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
surfnor= obr->vlaknodes[nr].surfnor= MEM_callocN(256*RE_SURFNOR_ELEMS*sizeof(float), "surfnor table");
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
else
return NULL;
}
return surfnor + (vlak->index & 255)*RE_SURFNOR_ELEMS;
}
Normal map tangents are now not always averaged at vertices anymore, but only when the UV's are connected. That fixes some artifacts when baking and using tangent space normal maps. It does mean increased memory usage because it now stores 4 tangents per face like UV's, and increased processing time, but there's no simple way around that.
2008-03-07 15:38:56 +00:00
float *RE_vlakren_get_nmap_tangent(ObjectRen *obr, VlakRen *vlak, int verify)
{
float *tangent;
int nr= vlak->index>>8;
tangent= obr->vlaknodes[nr].tangent;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (tangent==NULL) {
if (verify)
Normal map tangents are now not always averaged at vertices anymore, but only when the UV's are connected. That fixes some artifacts when baking and using tangent space normal maps. It does mean increased memory usage because it now stores 4 tangents per face like UV's, and increased processing time, but there's no simple way around that.
2008-03-07 15:38:56 +00:00
tangent= obr->vlaknodes[nr].tangent= MEM_callocN(256*RE_NMAP_TANGENT_ELEMS*sizeof(float), "tangent table");
else
return NULL;
}
return tangent + (vlak->index & 255)*RE_NMAP_TANGENT_ELEMS;
}
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
RadFace **RE_vlakren_get_radface(ObjectRen *obr, VlakRen *vlak, int verify)
{
RadFace **radface;
int nr= vlak->index>>8;
radface= obr->vlaknodes[nr].radface;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (radface==NULL) {
if (verify)
style cleanup
2012-12-28 14:19:05 +00:00
radface = obr->vlaknodes[nr].radface= MEM_callocN(256 * RE_RADFACE_ELEMS * sizeof(void *), "radface table");
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
else
return NULL;
}
return radface + (vlak->index & 255)*RE_RADFACE_ELEMS;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
VlakRen *RE_vlakren_copy(ObjectRen *obr, VlakRen *vlr)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
{
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
VlakRen *vlr1 = RE_findOrAddVlak(obr, obr->totvlak++);
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
MTFace *mtface, *mtface1;
MCol *mcol, *mcol1;
Normal map tangents are now not always averaged at vertices anymore, but only when the UV's are connected. That fixes some artifacts when baking and using tangent space normal maps. It does mean increased memory usage because it now stores 4 tangents per face like UV's, and increased processing time, but there's no simple way around that.
2008-03-07 15:38:56 +00:00
float *surfnor, *surfnor1, *tangent, *tangent1;
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
int *origindex, *origindex1;
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
RadFace **radface, **radface1;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
int i, index = vlr1->index;
char *name;
*vlr1= *vlr;
vlr1->index= index;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
for (i=0; (mtface=RE_vlakren_get_tface(obr, vlr, i, &name, 0)) != NULL; i++) {
mtface1= RE_vlakren_get_tface(obr, vlr1, i, &name, 1);
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
memcpy(mtface1, mtface, sizeof(MTFace)*RE_MTFACE_ELEMS);
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
for (i=0; (mcol=RE_vlakren_get_mcol(obr, vlr, i, &name, 0)) != NULL; i++) {
mcol1= RE_vlakren_get_mcol(obr, vlr1, i, &name, 1);
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
memcpy(mcol1, mcol, sizeof(MCol)*RE_MCOL_ELEMS);
}
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
origindex= RE_vlakren_get_origindex(obr, vlr, 0);
style cleanup: also correct doc example for 'foreach_get/set'
2013-02-02 00:34:34 +00:00
if (origindex) {
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
origindex1= RE_vlakren_get_origindex(obr, vlr1, 1);
/* Just an int, but memcpy for consistency. */
memcpy(origindex1, origindex, sizeof(int)*RE_VLAK_ORIGINDEX_ELEMS);
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
surfnor= RE_vlakren_get_surfnor(obr, vlr, 0);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (surfnor) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
surfnor1= RE_vlakren_get_surfnor(obr, vlr1, 1);
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
copy_v3_v3(surfnor1, surfnor);
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
Normal map tangents are now not always averaged at vertices anymore, but only when the UV's are connected. That fixes some artifacts when baking and using tangent space normal maps. It does mean increased memory usage because it now stores 4 tangents per face like UV's, and increased processing time, but there's no simple way around that.
2008-03-07 15:38:56 +00:00
tangent= RE_vlakren_get_nmap_tangent(obr, vlr, 0);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (tangent) {
Normal map tangents are now not always averaged at vertices anymore, but only when the UV's are connected. That fixes some artifacts when baking and using tangent space normal maps. It does mean increased memory usage because it now stores 4 tangents per face like UV's, and increased processing time, but there's no simple way around that.
2008-03-07 15:38:56 +00:00
tangent1= RE_vlakren_get_nmap_tangent(obr, vlr1, 1);
memcpy(tangent1, tangent, sizeof(float)*RE_NMAP_TANGENT_ELEMS);
}
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
radface= RE_vlakren_get_radface(obr, vlr, 0);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (radface) {
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
radface1= RE_vlakren_get_radface(obr, vlr1, 1);
*radface1= *radface;
}
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
return vlr1;
}
code cleanup: use const float and define array size
2012-05-26 22:21:56 +00:00
void RE_vlakren_get_normal(Render *UNUSED(re), ObjectInstanceRen *obi, VlakRen *vlr, float r_nor[3])
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
{
2.5: remove vertex normal flip option, this is more harmful than helpful in many cases, and also gave incorrect rim lighting.
2010-07-30 11:50:00 +00:00
float (*nmat)[3]= obi->nmat;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (obi->flag & R_TRANSFORMED) {
code cleanup: use const float and define array size
2012-05-26 22:21:56 +00:00
mul_v3_m3v3(r_nor, nmat, vlr->n);
normalize_v3(r_nor);
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
}
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
else {
code cleanup: use const float and define array size
2012-05-26 22:21:56 +00:00
copy_v3_v3(r_nor, vlr->n);
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
}
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
void RE_set_customdata_names(ObjectRen *obr, CustomData *data)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
{
/* CustomData layer names are stored per object here, because the
style cleanup: comment blocks
2012-03-09 18:28:30 +00:00
* DerivedMesh which stores the layers is freed */
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
CustomDataLayer *layer;
Fix for bug #17580: crash rendering instanced objects with halo material attach to the object instead of the mesh. Also for bug #13489: avoid a crash rendering with invalid active vcol layer, most likely caused by a bug that was already fixed.
2008-09-09 14:16:54 +00:00
int numtf = 0, numcol = 0, i, mtfn, mcn;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
if (CustomData_has_layer(data, CD_MTFACE)) {
Fix for bug #17580: crash rendering instanced objects with halo material attach to the object instead of the mesh. Also for bug #13489: avoid a crash rendering with invalid active vcol layer, most likely caused by a bug that was already fixed.
2008-09-09 14:16:54 +00:00
numtf= CustomData_number_of_layers(data, CD_MTFACE);
obr->mtface= MEM_callocN(sizeof(*obr->mtface)*numtf, "mtfacenames");
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
if (CustomData_has_layer(data, CD_MCOL)) {
Fix for bug #17580: crash rendering instanced objects with halo material attach to the object instead of the mesh. Also for bug #13489: avoid a crash rendering with invalid active vcol layer, most likely caused by a bug that was already fixed.
2008-09-09 14:16:54 +00:00
numcol= CustomData_number_of_layers(data, CD_MCOL);
obr->mcol= MEM_callocN(sizeof(*obr->mcol)*numcol, "mcolnames");
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
for (i=0, mtfn=0, mcn=0; i < data->totlayer; i++) {
layer= &data->layers[i];
if (layer->type == CD_MTFACE) {
strcpy() --> BLI_strncpy(), where source strings are not fixed and target size is known.
2011-10-19 23:10:54 +00:00
BLI_strncpy(obr->mtface[mtfn++], layer->name, sizeof(layer->name));
Fix for bug #17580: crash rendering instanced objects with halo material attach to the object instead of the mesh. Also for bug #13489: avoid a crash rendering with invalid active vcol layer, most likely caused by a bug that was already fixed.
2008-09-09 14:16:54 +00:00
obr->actmtface= CLAMPIS(layer->active_rnd, 0, numtf);
Fix for bug #11650: removing uv layers in editmode did not work correct. Fix for bug #11661: bake to viewport uv layer instead of render uv layer.
2008-05-15 19:40:09 +00:00
obr->bakemtface= layer->active;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
else if (layer->type == CD_MCOL) {
strcpy() --> BLI_strncpy(), where source strings are not fixed and target size is known.
2011-10-19 23:10:54 +00:00
BLI_strncpy(obr->mcol[mcn++], layer->name, sizeof(layer->name));
Fix for bug #17580: crash rendering instanced objects with halo material attach to the object instead of the mesh. Also for bug #13489: avoid a crash rendering with invalid active vcol layer, most likely caused by a bug that was already fixed.
2008-09-09 14:16:54 +00:00
obr->actmcol= CLAMPIS(layer->active_rnd, 0, numcol);
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
}
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
VlakRen *RE_findOrAddVlak(ObjectRen *obr, int nr)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
{
VlakTableNode *temp;
VlakRen *v;
int a;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (nr<0) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
printf("error in findOrAddVlak: %d\n", nr);
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
return obr->vlaknodes[0].vlak;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
a= nr>>8;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (a>=obr->vlaknodeslen-1) { /* Need to allocate more columns..., and keep last element NULL for free loop */
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
temp= obr->vlaknodes;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
obr->vlaknodes= MEM_mallocN(sizeof(VlakTableNode)*(obr->vlaknodeslen+TABLEINITSIZE), "vlaknodes");
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (temp) memcpy(obr->vlaknodes, temp, obr->vlaknodeslen*sizeof(VlakTableNode));
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
memset(obr->vlaknodes+obr->vlaknodeslen, 0, TABLEINITSIZE*sizeof(VlakTableNode));
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->vlaknodeslen+=TABLEINITSIZE; /*Does this really need to be power of 2?*/
style cleanup: trailing tabs & expand some non prefix tabs into spaces.
2012-10-21 05:46:41 +00:00
if (temp) MEM_freeN(temp);
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
v= obr->vlaknodes[a].vlak;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (v==NULL) {
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
int i;
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
v= (VlakRen *)MEM_callocN(256*sizeof(VlakRen), "findOrAddVlak");
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->vlaknodes[a].vlak= v;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (i= (nr & 0xFFFFFF00), a=0; a<256; a++, i++)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
v[a].index= i;
}
v+= (nr & 255);
return v;
}
/* ------------------------------------------------------------------------ */
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
float *RE_strandren_get_surfnor(ObjectRen *obr, StrandRen *strand, int verify)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
{
float *surfnor;
int nr= strand->index>>8;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
surfnor= obr->strandnodes[nr].surfnor;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (surfnor==NULL) {
if (verify)
- Removed the Save Buffers/FSA restriction that the width and height must be divisible by the number of parts. - OpenEXR code also has some changes to allow writing mipmapped images, but that is unused at the moment.
2008-02-06 13:45:07 +00:00
surfnor= obr->strandnodes[nr].surfnor= MEM_callocN(256*RE_SURFNOR_ELEMS*sizeof(float), "surfnor strand table");
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
else
return NULL;
}
return surfnor + (strand->index & 255)*RE_SURFNOR_ELEMS;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
float *RE_strandren_get_uv(ObjectRen *obr, StrandRen *strand, int n, char **name, int verify)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
{
StrandTableNode *node;
int nr= strand->index>>8, strandindex= (strand->index&255);
int index= (n<<8) + strandindex;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
node= &obr->strandnodes[nr];
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (verify) {
if (n>=node->totuv) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
float *uv= node->uv;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
int size= (n+1)*256;
- Removed the Save Buffers/FSA restriction that the width and height must be divisible by the number of parts. - OpenEXR code also has some changes to allow writing mipmapped images, but that is unused at the moment.
2008-02-06 13:45:07 +00:00
node->uv= MEM_callocN(size*sizeof(float)*RE_UV_ELEMS, "strand uv table");
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (uv) {
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
size= node->totuv*256;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
memcpy(node->uv, uv, size*sizeof(float)*RE_UV_ELEMS);
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
MEM_freeN(uv);
}
node->totuv= n+1;
}
}
else {
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (n>=node->totuv)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
return NULL;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (name) *name= obr->mtface[n];
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
return node->uv + index*RE_UV_ELEMS;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MCol *RE_strandren_get_mcol(ObjectRen *obr, StrandRen *strand, int n, char **name, int verify)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
{
StrandTableNode *node;
int nr= strand->index>>8, strandindex= (strand->index&255);
int index= (n<<8) + strandindex;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
node= &obr->strandnodes[nr];
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (verify) {
if (n>=node->totmcol) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MCol *mcol= node->mcol;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
int size= (n+1)*256;
- Removed the Save Buffers/FSA restriction that the width and height must be divisible by the number of parts. - OpenEXR code also has some changes to allow writing mipmapped images, but that is unused at the moment.
2008-02-06 13:45:07 +00:00
node->mcol= MEM_callocN(size*sizeof(MCol)*RE_MCOL_ELEMS, "strand mcol table");
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mcol) {
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
size= node->totmcol*256;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
memcpy(node->mcol, mcol, size*sizeof(MCol)*RE_MCOL_ELEMS);
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
MEM_freeN(mcol);
}
node->totmcol= n+1;
}
}
else {
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (n>=node->totmcol)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
return NULL;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (name) *name= obr->mcol[n];
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
return node->mcol + index*RE_MCOL_ELEMS;
}
Strand Render Simplification ============================ - Strand render now has options to remove child strands as the object's faces becomes smaller, in the Simplification particle panel. - "Reference Size" is the approximate size of the object on screen, after which simplification starts. - "Rate" is how fast strands are removed. - "Transition" is the percentage of strands being faded out as they are removed. - Another "Viewport" option removes strands on faces that are outside of the viewport. "Rate" again controls how fast these are removed. - Strand render in Blender Units now has an adjustable minimum width. Below this minimum width, strands start fading out instead of getting smaller.
2007-12-20 16:35:27 +00:00
float *RE_strandren_get_simplify(struct ObjectRen *obr, struct StrandRen *strand, int verify)
{
float *simplify;
int nr= strand->index>>8;
simplify= obr->strandnodes[nr].simplify;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (simplify==NULL) {
if (verify)
- Removed the Save Buffers/FSA restriction that the width and height must be divisible by the number of parts. - OpenEXR code also has some changes to allow writing mipmapped images, but that is unused at the moment.
2008-02-06 13:45:07 +00:00
simplify= obr->strandnodes[nr].simplify= MEM_callocN(256*RE_SIMPLIFY_ELEMS*sizeof(float), "simplify strand table");
Strand Render Simplification ============================ - Strand render now has options to remove child strands as the object's faces becomes smaller, in the Simplification particle panel. - "Reference Size" is the approximate size of the object on screen, after which simplification starts. - "Rate" is how fast strands are removed. - "Transition" is the percentage of strands being faded out as they are removed. - Another "Viewport" option removes strands on faces that are outside of the viewport. "Rate" again controls how fast these are removed. - Strand render in Blender Units now has an adjustable minimum width. Below this minimum width, strands start fading out instead of getting smaller.
2007-12-20 16:35:27 +00:00
else
return NULL;
}
return simplify + (strand->index & 255)*RE_SIMPLIFY_ELEMS;
}
Approximate Ambient Occlusion ============================= A new approximate ambient occlusion method has been added, next to the existing one based on raytracing. This method is specifically targetted at use in animations, since it is inherently noise free, and so will not flicker across frames. http://www.blender.org/development/current-projects/changes-since-244/approximate-ambient-occlusion/ http://peach.blender.org/index.php/approximate-ambient-occlusion/ Further improvements are still needed, but it can be tested already. There are still a number of known issues: - Bias errors on backfaces. - For performance, instanced object do not occlude currently. - Sky textures don't work well, the derivatives for texture evaluation are not correct. - Multiple passes do not work entirely correct (they are not accurate to begin with, but could be better).
2008-01-17 19:27:16 +00:00
int *RE_strandren_get_face(ObjectRen *obr, StrandRen *strand, int verify)
{
int *face;
int nr= strand->index>>8;
face= obr->strandnodes[nr].face;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (face==NULL) {
if (verify)
- Removed the Save Buffers/FSA restriction that the width and height must be divisible by the number of parts. - OpenEXR code also has some changes to allow writing mipmapped images, but that is unused at the moment.
2008-02-06 13:45:07 +00:00
face= obr->strandnodes[nr].face= MEM_callocN(256*RE_FACE_ELEMS*sizeof(int), "face strand table");
Approximate Ambient Occlusion ============================= A new approximate ambient occlusion method has been added, next to the existing one based on raytracing. This method is specifically targetted at use in animations, since it is inherently noise free, and so will not flicker across frames. http://www.blender.org/development/current-projects/changes-since-244/approximate-ambient-occlusion/ http://peach.blender.org/index.php/approximate-ambient-occlusion/ Further improvements are still needed, but it can be tested already. There are still a number of known issues: - Bias errors on backfaces. - For performance, instanced object do not occlude currently. - Sky textures don't work well, the derivatives for texture evaluation are not correct. - Multiple passes do not work entirely correct (they are not accurate to begin with, but could be better).
2008-01-17 19:27:16 +00:00
else
return NULL;
}
return face + (strand->index & 255)*RE_FACE_ELEMS;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
/* winspeed is exception, it is stored per instance */
float *RE_strandren_get_winspeed(ObjectInstanceRen *obi, StrandRen *strand, int verify)
{
float *winspeed;
int totvector;
winspeed= obi->vectors;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (winspeed==NULL) {
if (verify) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
totvector= obi->obr->totvert + obi->obr->totstrand;
- Removed the Save Buffers/FSA restriction that the width and height must be divisible by the number of parts. - OpenEXR code also has some changes to allow writing mipmapped images, but that is unused at the moment.
2008-02-06 13:45:07 +00:00
winspeed= obi->vectors= MEM_callocN(totvector*RE_WINSPEED_ELEMS*sizeof(float), "winspeed strand table");
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
}
else
return NULL;
}
return winspeed + (obi->obr->totvert + strand->index)*RE_WINSPEED_ELEMS;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
StrandRen *RE_findOrAddStrand(ObjectRen *obr, int nr)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
{
StrandTableNode *temp;
StrandRen *v;
int a;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (nr<0) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
printf("error in findOrAddStrand: %d\n", nr);
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
return obr->strandnodes[0].strand;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
a= nr>>8;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (a>=obr->strandnodeslen-1) { /* Need to allocate more columns..., and keep last element NULL for free loop */
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
temp= obr->strandnodes;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
obr->strandnodes= MEM_mallocN(sizeof(StrandTableNode)*(obr->strandnodeslen+TABLEINITSIZE), "strandnodes");
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (temp) memcpy(obr->strandnodes, temp, obr->strandnodeslen*sizeof(StrandTableNode));
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
memset(obr->strandnodes+obr->strandnodeslen, 0, TABLEINITSIZE*sizeof(StrandTableNode));
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->strandnodeslen+=TABLEINITSIZE; /*Does this really need to be power of 2?*/
style cleanup: trailing tabs & expand some non prefix tabs into spaces.
2012-10-21 05:46:41 +00:00
if (temp) MEM_freeN(temp);
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
v= obr->strandnodes[a].strand;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (v==NULL) {
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
int i;
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
v= (StrandRen *)MEM_callocN(256*sizeof(StrandRen), "findOrAddStrand");
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->strandnodes[a].strand= v;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (i= (nr & 0xFFFFFF00), a=0; a<256; a++, i++)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
v[a].index= i;
}
v+= (nr & 255);
return v;
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
StrandBuffer *RE_addStrandBuffer(ObjectRen *obr, int totvert)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
{
StrandBuffer *strandbuf;
strandbuf= MEM_callocN(sizeof(StrandBuffer), "StrandBuffer");
strandbuf->vert= MEM_callocN(sizeof(StrandVert)*totvert, "StrandVert");
strandbuf->totvert= totvert;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
strandbuf->obr= obr;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
Bounding box clipping in the render engine. Now bounding boxes are computed per object, and checked first before zbuffering objects. For strands, bounding boxes are computed per original face in the mesh. Overall the speed improvement from this is quite small (zbuffering is rarely the bottleneck), but it seems a sensible thing to do anyway.
2008-01-29 12:20:42 +00:00
obr->strandbuf= strandbuf;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
return strandbuf;
}
/* ------------------------------------------------------------------------ */
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
ObjectRen *RE_addRenderObject(Render *re, Object *ob, Object *par, int index, int psysindex, int lay)
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
{
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
ObjectRen *obr= MEM_callocN(sizeof(ObjectRen), "object render struct");
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
BLI_addtail(&re->objecttable, obr);
obr->ob= ob;
obr->par= par;
obr->index= index;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->psysindex= psysindex;
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
obr->lay= lay;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
return obr;
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
}
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
void free_renderdata_vertnodes(VertTableNode *vertnodes)
{
int a;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vertnodes==NULL) return;
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (a=0; vertnodes[a].vert; a++) {
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
MEM_freeN(vertnodes[a].vert);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vertnodes[a].rad)
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
MEM_freeN(vertnodes[a].rad);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vertnodes[a].strand)
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
MEM_freeN(vertnodes[a].strand);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vertnodes[a].tangent)
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
MEM_freeN(vertnodes[a].tangent);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vertnodes[a].stress)
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
MEM_freeN(vertnodes[a].stress);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vertnodes[a].winspeed)
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
MEM_freeN(vertnodes[a].winspeed);
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
if (vertnodes[a].origindex)
MEM_freeN(vertnodes[a].origindex);
Three features; - Live scanline updates while rendering Using a timer system, each second now the tiles that are being processed are checked if they could use display. To make this work pretty, I had to use the threaded 'tile processor' for a single thread too, but that's now proven to be stable. Also note that these updates draw per layer, including ztransp progress separately from solid render. - Recode of ztransp OSA Until now (since blender 1.0) the ztransp part was fully rendered and added on top of the solid part with alpha-over. This adding was done before the solid part applied sub-pixel sample filtering, causing the ztransp layer to be always too blurry. Now the ztransp layer uses same sub=pixel filter, resulting in the same AA level (and filter results) as the solid part. Quite noticable with hair renders. - Vector buffer support & preliminary vector-blur Node Using the "Render Layer" panel "Vector" pass button, the motion vectors per pixel are calculated and stored. Accessible via the Compositor. The vector-blur node is horrible btw! It just uses the length of the vector to apply a filter like with current (z)blur. I'm committing it anyway, I'll experiment with it further, and who knows some surprise code shows up!
2006-01-31 21:49:05 +00:00
}
MEM_freeN(vertnodes);
}
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
void free_renderdata_vlaknodes(VlakTableNode *vlaknodes)
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
{
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
int a;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vlaknodes==NULL) return;
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (a=0; vlaknodes[a].vlak; a++) {
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
MEM_freeN(vlaknodes[a].vlak);
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vlaknodes[a].mtface)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
MEM_freeN(vlaknodes[a].mtface);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vlaknodes[a].mcol)
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
MEM_freeN(vlaknodes[a].mcol);
style cleanup: also correct doc example for 'foreach_get/set'
2013-02-02 00:34:34 +00:00
if (vlaknodes[a].origindex)
Patch [#29035] Vertex colour baking There is a new option in the Bake panel to enable baking to vertex colors. Unlike regular baking, this mode does not require a UV map or image to bake to, however the object must have a vertex color layer. Thanks to: - AutoCRC for funding - Brech van Lommel and Dalai Felinto for their initial advice on how to implement it - Campbell Barton for helping to make this feature work with modifiers and bmesh
2013-01-02 00:05:30 +00:00
MEM_freeN(vlaknodes[a].origindex);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vlaknodes[a].surfnor)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
MEM_freeN(vlaknodes[a].surfnor);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vlaknodes[a].tangent)
Normal map tangents are now not always averaged at vertices anymore, but only when the UV's are connected. That fixes some artifacts when baking and using tangent space normal maps. It does mean increased memory usage because it now stores 4 tangents per face like UV's, and increased processing time, but there's no simple way around that.
2008-03-07 15:38:56 +00:00
MEM_freeN(vlaknodes[a].tangent);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vlaknodes[a].radface)
Minor render memory usage optimization, removed layer and radface from VlakRen, saves about 100mb for 10 million faces.
2008-01-23 13:35:51 +00:00
MEM_freeN(vlaknodes[a].radface);
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
}
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
MEM_freeN(vlaknodes);
}
minor cleanup: make functions static, use NULL for pointer comparisons, also fixed a possible bug assigning incorrect DPX function types to imbuf.
2011-05-10 13:11:36 +00:00
static void free_renderdata_strandnodes(StrandTableNode *strandnodes)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
{
int a;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strandnodes==NULL) return;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (a=0; strandnodes[a].strand; a++) {
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
MEM_freeN(strandnodes[a].strand);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strandnodes[a].uv)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
MEM_freeN(strandnodes[a].uv);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strandnodes[a].mcol)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
MEM_freeN(strandnodes[a].mcol);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strandnodes[a].winspeed)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
MEM_freeN(strandnodes[a].winspeed);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strandnodes[a].surfnor)
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
MEM_freeN(strandnodes[a].surfnor);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strandnodes[a].simplify)
Strand Render Simplification ============================ - Strand render now has options to remove child strands as the object's faces becomes smaller, in the Simplification particle panel. - "Reference Size" is the approximate size of the object on screen, after which simplification starts. - "Rate" is how fast strands are removed. - "Transition" is the percentage of strands being faded out as they are removed. - Another "Viewport" option removes strands on faces that are outside of the viewport. "Rate" again controls how fast these are removed. - Strand render in Blender Units now has an adjustable minimum width. Below this minimum width, strands start fading out instead of getting smaller.
2007-12-20 16:35:27 +00:00
MEM_freeN(strandnodes[a].simplify);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strandnodes[a].face)
Approximate Ambient Occlusion ============================= A new approximate ambient occlusion method has been added, next to the existing one based on raytracing. This method is specifically targetted at use in animations, since it is inherently noise free, and so will not flicker across frames. http://www.blender.org/development/current-projects/changes-since-244/approximate-ambient-occlusion/ http://peach.blender.org/index.php/approximate-ambient-occlusion/ Further improvements are still needed, but it can be tested already. There are still a number of known issues: - Bias errors on backfaces. - For performance, instanced object do not occlude currently. - Sky textures don't work well, the derivatives for texture evaluation are not correct. - Multiple passes do not work entirely correct (they are not accurate to begin with, but could be better).
2008-01-17 19:27:16 +00:00
MEM_freeN(strandnodes[a].face);
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
MEM_freeN(strandnodes);
}
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
void free_renderdata_tables(Render *re)
{
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
ObjectInstanceRen *obi;
ObjectRen *obr;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
StrandBuffer *strandbuf;
int a=0;
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (obr=re->objecttable.first; obr; obr=obr->next) {
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->vertnodes) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
free_renderdata_vertnodes(obr->vertnodes);
obr->vertnodes= NULL;
obr->vertnodeslen= 0;
}
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->vlaknodes) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
free_renderdata_vlaknodes(obr->vlaknodes);
obr->vlaknodes= NULL;
obr->vlaknodeslen= 0;
obr->totvlak= 0;
}
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->bloha) {
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (a=0; obr->bloha[a]; a++)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MEM_freeN(obr->bloha[a]);
MEM_freeN(obr->bloha);
obr->bloha= NULL;
obr->blohalen= 0;
}
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->strandnodes) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
free_renderdata_strandnodes(obr->strandnodes);
obr->strandnodes= NULL;
obr->strandnodeslen= 0;
}
Bounding box clipping in the render engine. Now bounding boxes are computed per object, and checked first before zbuffering objects. For strands, bounding boxes are computed per original face in the mesh. Overall the speed improvement from this is quite small (zbuffering is rarely the bottleneck), but it seems a sensible thing to do anyway.
2008-01-29 12:20:42 +00:00
strandbuf= obr->strandbuf;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (strandbuf) {
if (strandbuf->vert) MEM_freeN(strandbuf->vert);
if (strandbuf->bound) MEM_freeN(strandbuf->bound);
Bounding box clipping in the render engine. Now bounding boxes are computed per object, and checked first before zbuffering objects. For strands, bounding boxes are computed per original face in the mesh. Overall the speed improvement from this is quite small (zbuffering is rarely the bottleneck), but it seems a sensible thing to do anyway.
2008-01-29 12:20:42 +00:00
MEM_freeN(strandbuf);
}
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->mtface)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MEM_freeN(obr->mtface);
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->mcol)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MEM_freeN(obr->mcol);
*Instance support at dupliverts/faces *Octree works nicely on hierarchic trees *the old code was quite destructive at the Isect variable changing isec->start, end, vec..now it only changes isec->labda (and hit results) Currently rendering a BVH of all objects, where each object has it own octree.
2009-06-11 16:44:01 +00:00
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->rayfaces) {
*Instance support at dupliverts/faces *Octree works nicely on hierarchic trees *the old code was quite destructive at the Isect variable changing isec->start, end, vec..now it only changes isec->labda (and hit results) Currently rendering a BVH of all objects, where each object has it own octree.
2009-06-11 16:44:01 +00:00
MEM_freeN(obr->rayfaces);
obr->rayfaces = NULL;
}
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->rayprimitives) {
*Added VlakPrimitive (this rayobject rimitive only stores ObjectRenderInstance and VlakRen pointers) - it difers from RayFace that localy stored the vertex coordinates. - basicaly this reduces memory usage
2009-09-24 22:55:57 +00:00
MEM_freeN(obr->rayprimitives);
obr->rayprimitives = NULL;
}
style cleanup, brackets in else/if, some indentation.
2012-03-06 18:40:15 +00:00
if (obr->raytree) {
*Instance support at dupliverts/faces *Octree works nicely on hierarchic trees *the old code was quite destructive at the Isect variable changing isec->start, end, vec..now it only changes isec->labda (and hit results) Currently rendering a BVH of all objects, where each object has it own octree.
2009-06-11 16:44:01 +00:00
RE_rayobject_free(obr->raytree);
obr->raytree = NULL;
}
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (re->objectinstance) {
style cleanup: changes to brace placement / newlines - for/while/if/switch
2012-04-28 06:31:57 +00:00
for (obi=re->instancetable.first; obi; obi=obi->next) {
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (obi->vectors)
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MEM_freeN(obi->vectors);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (obi->raytree)
*Instance support at dupliverts/faces *Octree works nicely on hierarchic trees *the old code was quite destructive at the Isect variable changing isec->start, end, vec..now it only changes isec->labda (and hit results) Currently rendering a BVH of all objects, where each object has it own octree.
2009-06-11 16:44:01 +00:00
RE_rayobject_free(obi->raytree);
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MEM_freeN(re->objectinstance);
re->objectinstance= NULL;
re->totinstance= 0;
re->instancetable.first= re->instancetable.last= NULL;
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
}
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (re->sortedhalos) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
MEM_freeN(re->sortedhalos);
re->sortedhalos= NULL;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
}
Added names to UV and vertex color layers, and display them as a list. Added support for multiple UVs in the render engine. This also involved changing the way faces are stored, to allow data to be added optionally per 256 faces, same as the existing system for vertices. A UV layer can be specified in the Map Input panel and the Geometry node by name. Leaving this field blank will default to the active UV layer. Also added sharing of face selection and hiding between UV layers, and at the same time improved syncing with editmode selection and hiding. Still to do: - Multi UV support for fastshade. - Multires and NMesh preservation of multiple UV sets.
2006-12-21 13:47:27 +00:00
BLI_freelistN(&re->customdata_names);
iImage based Vector Blur After a couple of experiments with variable blur filters, I tried a more interesting, and who knows... original approach. :) First watch results here: http://www.blender.org/bf/rt0001_0030.avi http://www.blender.org/bf/hand0001_0060.avi These are the steps in producing such results: - In preprocess, the speed vectors to previous and next frame are calculated. Speed vectors are screen-aligned and in pixel size. - while rendering, these vectors get calculated per sample, and accumulated in the vector buffer checking for "minimum speed". (on start the vector buffer is initialized on max speed). - After render: - The entire image, all pixels, then is converted to quad polygons. - Also the z value of the pixels is assigned to the polygons - The vertices for the quads use averaged speed vectors (of the 4 corner faces), using a 'minimum but non-zero' speed rule. This minimal speed trick works very well to prevent 'tearing' apart when multiple faces move in different directions in a pixel, or to be able to separate moving pixels clearly from non-moving ones - So, now we have a sort of 'mask' of quad polygons. The previous steps guaranteed that this mask doesn't have antialias color info, and has speed vectors that ensure individual parts to move nicely without tearing effects. The Z allows multiple layers of moving masks. - Then, in temporal buffer, faces get tagged if they move or not - These tags then go to an anti-alias routine, which assigns alpha values to edge faces, based on the method we used in past to antialias bitmaps (still in our code, check the antialias.c in imbuf!) - finally, the tag buffer is used to tag which z values of the original image have to be included (to allow blur go behind stuff). - OK, now we're ready for accumulating! In a loop, all faces then get drawn (with zbuffer) with increasing influence of their speed vectors. The resulting image then is accumulated on top of the original with a decreasing weighting value. It sounds all quite complex... but the speed is still encouraging. Above images have 64 mblur steps, which takes about 1-3 seconds per frame. Usage notes: - Make sure the render-layer has passes 'Vector' and 'Z' on. - add in Compositor the VectorBlur node, and connect the image, Z and speed to the inputs. - The node allows to set amount of steps (10 steps = 10 forward, 10 back). and to set a maximum speed in pixels... to prevent extreme moving things to blur too wide.
2006-02-06 22:11:50 +00:00
BLI_freelistN(&re->objecttable);
- Make dupliverts inside a dupligroup render and draw as instances. - Possible fix for a dupli render memory leak, but i'm not sure.
2008-01-04 17:48:07 +00:00
BLI_freelistN(&re->instancetable);
New feature: "Stress" texture input channel (As usual movies disappears after while) Face example showing stress values on a blend. White is stretch, black is squeeze http://www.blender.org/bf/0001_0014.avi Quick test with softbody stretch http://www.blender.org/bf/0001_0100.avi Based on the difference of the "Orco" (original undeformed coordinate) and the actual render coordinate, a stress value is computed to make textures react to stretching or wrinking skin. The texture coordinate is neutral (0) on relaxed state. -1 is squeezed to zero, +1 is stretched to infinity. Note that scaling (object itself or parent) also will result in stress values. The reason for the huge commit is a cleanup in allocating memory for the vertices. These were growing too large with new options, so now it allocates the optional coordinates dynamically. Saves about 20 MB memory per 1M vertices already. But best of all is that I now can add much more fun... so tangents, here we come!
2005-12-08 22:05:42 +00:00
}
Initial revision
2002-10-12 11:37:38 +00:00
/* ------------------------------------------------------------------------ */
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
HaloRen *RE_findOrAddHalo(ObjectRen *obr, int nr)
Initial revision
2002-10-12 11:37:38 +00:00
{
Dynamic Face/Vert/Halo/Lamp tables: User Info: Hard coded limits on the total number of face, verts, halos, and lamps is gone. Blender now allocates the tables for these on an as needed basis. As long as your system can come up with the memory, you won't run out. As a bonus, it also uses slightly less memory on smaller scenes. Coder info: This has been in tuhopuu for a while, but I don't know how hard it has been tested. Since it now allocates only an initial 1024 tables (of 256 verts/faces/halos each), it seems like it has been put through it's paces. Lamps are allocated one at a time, and I start with 256. I rendered 2.5M Faces/Verts/Halos. 4444 lamps. None the less, I left a few printf's in the realocation to hunt bugs. I'll take them out just before the release freeze. Also, be on the lookout for other "sanity checks" that assume a limited number of the above items. I think I got them all, but you never know.
2003-10-19 21:47:03 +00:00
HaloRen *h, **temp;
Initial revision
2002-10-12 11:37:38 +00:00
int a;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (nr<0) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
printf("error in findOrAddHalo: %d\n", nr);
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
return NULL;
Initial revision
2002-10-12 11:37:38 +00:00
}
a= nr>>8;
Dynamic Face/Vert/Halo/Lamp tables: User Info: Hard coded limits on the total number of face, verts, halos, and lamps is gone. Blender now allocates the tables for these on an as needed basis. As long as your system can come up with the memory, you won't run out. As a bonus, it also uses slightly less memory on smaller scenes. Coder info: This has been in tuhopuu for a while, but I don't know how hard it has been tested. Since it now allocates only an initial 1024 tables (of 256 verts/faces/halos each), it seems like it has been put through it's paces. Lamps are allocated one at a time, and I start with 256. I rendered 2.5M Faces/Verts/Halos. 4444 lamps. None the less, I left a few printf's in the realocation to hunt bugs. I'll take them out just before the release freeze. Also, be on the lookout for other "sanity checks" that assume a limited number of the above items. I think I got them all, but you never know.
2003-10-19 21:47:03 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (a>=obr->blohalen-1) { /* Need to allocate more columns..., and keep last element NULL for free loop */
Solved AO smooth rendering, by introducing a "Bias" value for smoothed rendered faces. http://www.blender3d.org/cms/Ambient_Occlusion.231.0.html ALso removed prints when allocating new 'face groups' etc, which wasn't informative at all.
2004-04-26 14:17:48 +00:00
//printf("Allocating %i more halo groups. %i total.\n",
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
// TABLEINITSIZE, obr->blohalen+TABLEINITSIZE );
temp=obr->bloha;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
style cleanup
2013-03-18 11:44:56 +00:00
obr->bloha = (HaloRen **)MEM_callocN(sizeof(void *) * (obr->blohalen + TABLEINITSIZE), "Bloha");
style cleanup
2012-12-28 14:19:05 +00:00
if (temp) memcpy(obr->bloha, temp, obr->blohalen*sizeof(void *));
memset(&(obr->bloha[obr->blohalen]), 0, TABLEINITSIZE * sizeof(void *));
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->blohalen+=TABLEINITSIZE; /*Does this really need to be power of 2?*/
style cleanup: trailing tabs & expand some non prefix tabs into spaces.
2012-10-21 05:46:41 +00:00
if (temp) MEM_freeN(temp);
Dynamic Face/Vert/Halo/Lamp tables: User Info: Hard coded limits on the total number of face, verts, halos, and lamps is gone. Blender now allocates the tables for these on an as needed basis. As long as your system can come up with the memory, you won't run out. As a bonus, it also uses slightly less memory on smaller scenes. Coder info: This has been in tuhopuu for a while, but I don't know how hard it has been tested. Since it now allocates only an initial 1024 tables (of 256 verts/faces/halos each), it seems like it has been put through it's paces. Lamps are allocated one at a time, and I start with 256. I rendered 2.5M Faces/Verts/Halos. 4444 lamps. None the less, I left a few printf's in the realocation to hunt bugs. I'll take them out just before the release freeze. Also, be on the lookout for other "sanity checks" that assume a limited number of the above items. I think I got them all, but you never know.
2003-10-19 21:47:03 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
h= obr->bloha[a];
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (h==NULL) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
h= (HaloRen *)MEM_callocN(256*sizeof(HaloRen), "findOrAdHalo");
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obr->bloha[a]= h;
Initial revision
2002-10-12 11:37:38 +00:00
}
h+= (nr & 255);
return h;
}
/* ------------------------------------------------------------------------- */
code cleanup: use const float and define array size
2012-05-26 22:21:56 +00:00
HaloRen *RE_inithalo(Render *re, ObjectRen *obr, Material *ma,
const float vec[3], const float vec1[3],
const float *orco, float hasize, float vectsize, int seed)
Initial revision
2002-10-12 11:37:38 +00:00
{
HaloRen *har;
MTex *mtex;
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
float tin, tr, tg, tb, ta;
Initial revision
2002-10-12 11:37:38 +00:00
float xn, yn, zn, texvec[3], hoco[4], hoco1[4];
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (hasize==0.0f) return NULL;
Initial revision
2002-10-12 11:37:38 +00:00
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
projectverto(vec, re->winmat, hoco);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (hoco[3]==0.0f) return NULL;
if (vec1) {
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
projectverto(vec1, re->winmat, hoco1);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (hoco1[3]==0.0f) return NULL;
Initial revision
2002-10-12 11:37:38 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
har= RE_findOrAddHalo(obr, obr->tothalo++);
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
copy_v3_v3(har->co, vec);
Initial revision
2002-10-12 11:37:38 +00:00
har->hasize= hasize;
IRC report fix: - Halos were not sorted, for ages! Meaning that they were rendered in order of creation, instead of back to front. - Made ESC testing for halo render to check per halo, is a fast routine now anyway.
2006-07-03 11:53:52 +00:00
/* actual projectvert is done in function project_renderdata() because of parts/border/pano */
/* we do it here for sorting of halos */
zn= hoco[3];
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->xs= 0.5f*re->winx*(hoco[0]/zn);
har->ys= 0.5f*re->winy*(hoco[1]/zn);
IRC report fix: - Halos were not sorted, for ages! Meaning that they were rendered in order of creation, instead of back to front. - Made ESC testing for halo render to check per halo, is a fast routine now anyway.
2006-07-03 11:53:52 +00:00
har->zs= 0x7FFFFF*(hoco[2]/zn);
har->zBufDist = 0x7FFFFFFF*(hoco[2]/zn);
Initial revision
2002-10-12 11:37:38 +00:00
/* halovect */
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vec1) {
Initial revision
2002-10-12 11:37:38 +00:00
har->type |= HA_VECT;
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
xn= har->xs - 0.5f*re->winx*(hoco1[0]/hoco1[3]);
yn= har->ys - 0.5f*re->winy*(hoco1[1]/hoco1[3]);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (xn==0.0f || (xn==0.0f && yn==0.0f)) zn= 0.0f;
Initial revision
2002-10-12 11:37:38 +00:00
else zn= atan2(yn, xn);
har->sin= sin(zn);
har->cos= cos(zn);
Math Lib * Convert all code to use new functions. * Branch maintainers may want to skip this commit, and run this conversion script instead, if they use a lot of math functions in new code: http://www.pasteall.org/9052/python
2009-11-10 20:43:45 +00:00
zn= len_v3v3(vec1, vec);
Initial revision
2002-10-12 11:37:38 +00:00
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->hasize= vectsize*zn + (1.0f-vectsize)*hasize;
Initial revision
2002-10-12 11:37:38 +00:00
Math Lib * Convert all code to use new functions. * Branch maintainers may want to skip this commit, and run this conversion script instead, if they use a lot of math functions in new code: http://www.pasteall.org/9052/python
2009-11-10 20:43:45 +00:00
sub_v3_v3v3(har->no, vec, vec1);
normalize_v3(har->no);
Initial revision
2002-10-12 11:37:38 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (ma->mode & MA_HALO_XALPHA) har->type |= HA_XALPHA;
Initial revision
2002-10-12 11:37:38 +00:00
har->alfa= ma->alpha;
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
har->r= ma->r;
har->g= ma->g;
har->b= ma->b;
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->add= (255.0f*ma->add);
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
har->mat= ma;
Initial revision
2002-10-12 11:37:38 +00:00
har->hard= ma->har;
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
har->seed= seed % 256;
Initial revision
2002-10-12 11:37:38 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (ma->mode & MA_STAR) har->starpoints= ma->starc;
if (ma->mode & MA_HALO_LINES) har->linec= ma->linec;
if (ma->mode & MA_HALO_RINGS) har->ringc= ma->ringc;
if (ma->mode & MA_HALO_FLARE) har->flarec= ma->flarec;
Initial revision
2002-10-12 11:37:38 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (ma->mtex[0]) {
Initial revision
2002-10-12 11:37:38 +00:00
style cleanup: if();
2012-10-07 09:48:59 +00:00
if (ma->mode & MA_HALOTEX) {
har->tex = 1;
}
else if (har->mat->septex & (1 << 0)) {
/* only 1 level textures */
}
Initial revision
2002-10-12 11:37:38 +00:00
else {
mtex= ma->mtex[0];
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
copy_v3_v3(texvec, vec);
Initial revision
2002-10-12 11:37:38 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->texco & TEXCO_NORM) {
Initial revision
2002-10-12 11:37:38 +00:00
;
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (mtex->texco & TEXCO_OBJECT) {
Initial revision
2002-10-12 11:37:38 +00:00
/* texvec[0]+= imatbase->ivec[0]; */
/* texvec[1]+= imatbase->ivec[1]; */
/* texvec[2]+= imatbase->ivec[2]; */
Math Lib * Convert all code to use new functions. * Branch maintainers may want to skip this commit, and run this conversion script instead, if they use a lot of math functions in new code: http://www.pasteall.org/9052/python
2009-11-10 20:43:45 +00:00
/* mul_m3_v3(imatbase->imat, texvec); */
Initial revision
2002-10-12 11:37:38 +00:00
}
else {
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (orco) {
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
copy_v3_v3(texvec, orco);
Initial revision
2002-10-12 11:37:38 +00:00
}
}
Fix projection texture painting crash It was caused by own mistake by not noticing externtex is used not only by render engine. Now this function uses pool passed as argument rather than using R.pool.
2013-01-22 08:05:00 +00:00
externtex(mtex, texvec, &tin, &tr, &tg, &tb, &ta, 0, re->pool);
Initial revision
2002-10-12 11:37:38 +00:00
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
yn= tin*mtex->colfac;
Code cleanup: fix some clang static checker warnings.
2011-12-22 20:29:44 +00:00
//zn= tin*mtex->alphafac;
Initial revision
2002-10-12 11:37:38 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->mapto & MAP_COL) {
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
zn= 1.0f-yn;
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
har->r= (yn*tr+ zn*ma->r);
har->g= (yn*tg+ zn*ma->g);
har->b= (yn*tb+ zn*ma->b);
Initial revision
2002-10-12 11:37:38 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->texco & TEXCO_UV) {
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
har->alfa= tin;
Initial revision
2002-10-12 11:37:38 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->mapto & MAP_ALPHA)
2.50: svn merge https://svn.blender.org/svnroot/bf-blender/trunk/blender -r19323:HEAD Notes: * blenderbuttons and ICON_SNAP_PEEL_OBJECT were not merged.
2009-04-20 15:06:46 +00:00
har->alfa= tin;
Initial revision
2002-10-12 11:37:38 +00:00
}
}
Fixed render time regression in Blender Internal It was caused by image threading safe commit and it was noticeable only on really multi-core CPU (like dual-socket Xeon stations), was not visible on core i7 machine. The reason of slowdown was spinlock around image buffer referencing, which lead to lots of cores waiting for single core and using image buffer after it was referenced was not so much longer than doing reference itself. The most clear solution here seemed to be introducing Image Pool which will contain list of loaded and referenced image buffers, so all threads could skip lock if the pool is used for reading only. Lock only needed in cases when buffer for requested image user is missing in the pool. This lock will happen only once per image so overall amount of locks is much less that it was before. To operate with pool: - BKE_image_pool_new() creates new pool - BKE_image_pool_free() destroys pool and dereferences all image buffers which were loaded to it - BKE_image_pool_acquire_ibuf() returns image buffer for given image and user. Pool could be NULL and in this case fallback to BKE_image_acquire_ibuf will happen. This helps to avoid lots to if(poll) checks in image sampling code. - BKE_image_pool_release_ibuf releases image buffer. In fact, it will only do something if pool is NULL, in all other case it'll equal to DoNothing operation.
2013-01-21 08:49:42 +00:00
har->pool = re->pool;
Initial revision
2002-10-12 11:37:38 +00:00
return har;
}
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
code cleanup: use const float and define array size
2012-05-26 22:21:56 +00:00
HaloRen *RE_inithalo_particle(Render *re, ObjectRen *obr, DerivedMesh *dm, Material *ma,
const float vec[3], const float vec1[3],
const float *orco, const float *uvco, float hasize, float vectsize, int seed, const float pa_co[3])
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
{
HaloRen *har;
MTex *mtex;
float tin, tr, tg, tb, ta;
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
float xn, yn, zn, texvec[3], hoco[4], hoco1[4], in[3], tex[3], out[3];
Bug fix: Halos didn't use texture alpha
2010-12-07 12:29:51 +00:00
int i, hasrgb;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (hasize==0.0f) return NULL;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
projectverto(vec, re->winmat, hoco);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (hoco[3]==0.0f) return NULL;
if (vec1) {
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
projectverto(vec1, re->winmat, hoco1);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (hoco1[3]==0.0f) return NULL;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
}
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
har= RE_findOrAddHalo(obr, obr->tothalo++);
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
copy_v3_v3(har->co, vec);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
har->hasize= hasize;
/* actual projectvert is done in function project_renderdata() because of parts/border/pano */
/* we do it here for sorting of halos */
zn= hoco[3];
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->xs= 0.5f*re->winx*(hoco[0]/zn);
har->ys= 0.5f*re->winy*(hoco[1]/zn);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
har->zs= 0x7FFFFF*(hoco[2]/zn);
har->zBufDist = 0x7FFFFFFF*(hoco[2]/zn);
/* halovect */
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (vec1) {
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
har->type |= HA_VECT;
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
xn= har->xs - 0.5f*re->winx*(hoco1[0]/hoco1[3]);
yn= har->ys - 0.5f*re->winy*(hoco1[1]/hoco1[3]);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (xn==0.0f || (xn==0.0f && yn==0.0f)) zn= 0.0;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
else zn= atan2(yn, xn);
har->sin= sin(zn);
har->cos= cos(zn);
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
zn= len_v3v3(vec1, vec)*0.5f;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->hasize= vectsize*zn + (1.0f-vectsize)*hasize;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
Math Lib * Convert all code to use new functions. * Branch maintainers may want to skip this commit, and run this conversion script instead, if they use a lot of math functions in new code: http://www.pasteall.org/9052/python
2009-11-10 20:43:45 +00:00
sub_v3_v3v3(har->no, vec, vec1);
normalize_v3(har->no);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (ma->mode & MA_HALO_XALPHA) har->type |= HA_XALPHA;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
har->alfa= ma->alpha;
har->r= ma->r;
har->g= ma->g;
har->b= ma->b;
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->add= (255.0f*ma->add);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
har->mat= ma;
har->hard= ma->har;
har->seed= seed % 256;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (ma->mode & MA_STAR) har->starpoints= ma->starc;
if (ma->mode & MA_HALO_LINES) har->linec= ma->linec;
if (ma->mode & MA_HALO_RINGS) har->ringc= ma->ringc;
if (ma->mode & MA_HALO_FLARE) har->flarec= ma->flarec;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if ((ma->mode & MA_HALOTEX) && ma->mtex[0])
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
har->tex= 1;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (i=0; i<MAX_MTEX; i++)
if (ma->mtex[i] && (ma->septex & (1<<i))==0) {
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
mtex= ma->mtex[i];
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
copy_v3_v3(texvec, vec);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->texco & TEXCO_NORM) {
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
;
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (mtex->texco & TEXCO_OBJECT) {
if (mtex->object)
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
mul_m4_v3(mtex->object->imat_ren, texvec);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (mtex->texco & TEXCO_GLOB) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
copy_v3_v3(texvec, vec);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (mtex->texco & TEXCO_UV && uvco) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
int uv_index=CustomData_get_named_layer_index(&dm->faceData, CD_MTFACE, mtex->uvname);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (uv_index<0)
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
uv_index=CustomData_get_active_layer_index(&dm->faceData, CD_MTFACE);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
uv_index-=CustomData_get_layer_index(&dm->faceData, CD_MTFACE);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
texvec[0]=2.0f*uvco[2*uv_index]-1.0f;
texvec[1]=2.0f*uvco[2*uv_index+1]-1.0f;
texvec[2]=0.0f;
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (mtex->texco & TEXCO_PARTICLE) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
/* particle coordinates in range [0, 1] */
"Particle" texture coordinates for halo materials: * Particle age can now be used as the texture x-coordinate, and location in a particle trail as the y-coordinate. * This finally enables particles in 2.5 to change their color (or any other texturable material property) by their age. * In 2.4x this was accomplished with the "100 frames == particle age", but this was both non-intuitive and slow as the animation system had to be recalculated for every particle. * Currently these are 2d coordinates (age/lifetime == x-coordinate, trail particle index/number of trail particles == y-coordinate), but other particle properties or possibly even a user definable property can be added as coordinates in the future. * On the code side this uses the same coordinate definition number (for halo materials) as strand coordinates (for surface materials). This is also nice as they intuitively mean nearly the same thing, i.e. along strand or during particle life.
2010-12-07 12:58:25 +00:00
texvec[0] = 2.f * pa_co[0] - 1.f;
texvec[1] = 2.f * pa_co[1] - 1.f;
texvec[2] = pa_co[2];
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (orco) {
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
copy_v3_v3(texvec, orco);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
}
Fix projection texture painting crash It was caused by own mistake by not noticing externtex is used not only by render engine. Now this function uses pool passed as argument rather than using R.pool.
2013-01-22 08:05:00 +00:00
hasrgb = externtex(mtex, texvec, &tin, &tr, &tg, &tb, &ta, 0, re->pool);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
//yn= tin*mtex->colfac;
Texture stack influences are now all separate values, and negative mapped values now have their influence negated instead. Also a few RNA changes for TextureSlot. Bumped subversion for the version patch.
2009-10-01 17:15:23 +00:00
//zn= tin*mtex->alphafac;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->mapto & MAP_COL) {
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
tex[0]=tr;
tex[1]=tg;
tex[2]=tb;
out[0]=har->r;
out[1]=har->g;
out[2]=har->b;
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
texture_rgb_blend(in, tex, out, tin, mtex->colfac, mtex->blendtype);
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
// zn= 1.0-yn;
//har->r= (yn*tr+ zn*ma->r);
//har->g= (yn*tg+ zn*ma->g);
//har->b= (yn*tb+ zn*ma->b);
har->r= in[0];
har->g= in[1];
har->b= in[2];
}
Bug fix: Halos didn't use texture alpha
2010-12-07 12:29:51 +00:00
/* alpha returned, so let's use it instead of intensity */
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (hasrgb)
Bug fix: Halos didn't use texture alpha
2010-12-07 12:29:51 +00:00
tin = ta;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->mapto & MAP_ALPHA)
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
har->alfa = texture_value_blend(mtex->def_var, har->alfa, tin, mtex->alphafac, mtex->blendtype);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->mapto & MAP_HAR)
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
har->hard = 1.0f+126.0f*texture_value_blend(mtex->def_var, ((float)har->hard)/127.0f, tin, mtex->hardfac, mtex->blendtype);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->mapto & MAP_RAYMIRR)
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
har->hasize = 100.0f*texture_value_blend(mtex->def_var, har->hasize/100.0f, tin, mtex->raymirrfac, mtex->blendtype);
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mtex->mapto & MAP_TRANSLU) {
style cleanup: whitespace / commas
2012-04-29 15:47:02 +00:00
float add = texture_value_blend(mtex->def_var, (float)har->add/255.0f, tin, mtex->translfac, mtex->blendtype);
UI Cleanup: Halo materials and textures * Textures applied to halo materials showed influence option for normal particles. This was really confusing, and with the cleanup I revealed a couple of hidden features too! ** Particles actually allow for textures to change halo size and hardness, but my guess is that nobody knew since the names were wrong in the ui! ** I also added the option to change the "add" value with a texture, since it was just silly not to have it. * Halo material properties are also a bit cleaner now.
2010-12-08 20:10:59 +00:00
CLAMP(add, 0.f, 1.f);
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->add = 255.0f*add;
UI Cleanup: Halo materials and textures * Textures applied to halo materials showed influence option for normal particles. This was really confusing, and with the cleanup I revealed a couple of hidden features too! ** Particles actually allow for textures to change halo size and hardness, but my guess is that nobody knew since the names were wrong in the ui! ** I also added the option to change the "add" value with a texture, since it was just silly not to have it. * Halo material properties are also a bit cleaner now.
2010-12-08 20:10:59 +00:00
}
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
/* now what on earth is this good for?? */
style cleanup
2012-04-21 12:51:47 +00:00
//if (mtex->texco & 16) {
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
// har->alfa= tin;
//}
}
Fixed render time regression in Blender Internal It was caused by image threading safe commit and it was noticeable only on really multi-core CPU (like dual-socket Xeon stations), was not visible on core i7 machine. The reason of slowdown was spinlock around image buffer referencing, which lead to lots of cores waiting for single core and using image buffer after it was referenced was not so much longer than doing reference itself. The most clear solution here seemed to be introducing Image Pool which will contain list of loaded and referenced image buffers, so all threads could skip lock if the pool is used for reading only. Lock only needed in cases when buffer for requested image user is missing in the pool. This lock will happen only once per image so overall amount of locks is much less that it was before. To operate with pool: - BKE_image_pool_new() creates new pool - BKE_image_pool_free() destroys pool and dereferences all image buffers which were loaded to it - BKE_image_pool_acquire_ibuf() returns image buffer for given image and user. Pool could be NULL and in this case fallback to BKE_image_acquire_ibuf will happen. This helps to avoid lots to if(poll) checks in image sampling code. - BKE_image_pool_release_ibuf releases image buffer. In fact, it will only do something if pool is NULL, in all other case it'll equal to DoNothing operation.
2013-01-21 08:49:42 +00:00
har->pool = re->pool;
Particles ========= Merge of the famous particle patch by Janne Karhu, a full rewrite of the Blender particle system. This includes: - Emitter, Hair and Reactor particle types. - Newtonian, Keyed and Boids physics. - Various particle visualisation and rendering types. - Vertex group and texture control for various properties. - Interpolated child particles from parents. - Hair editing with combing, growing, cutting, .. . - Explode modifier. - Harmonic, Magnetic fields, and multiple falloff types. .. and lots of other things, some more info is here: http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite http://wiki.blender.org/index.php/BlenderDev/Particles_Rewrite_Doc The new particle system cannot be backwards compatible. Old particle systems are being converted to the new system, but will require tweaking to get them looking the same as before. Point Cache =========== The new system to replace manual baking, based on automatic caching on disk. This is currently used by softbodies and the particle system. See the Cache API section on: http://wiki.blender.org/index.php/BlenderDev/PhysicsSprint Documentation ============= These new features still need good docs for the release logs, help for this is appreciated.
2007-11-26 22:09:57 +00:00
return har;
}
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
/* -------------------------- operations on entire database ----------------------- */
/* ugly function for halos in panorama */
Code cleanup: remove TRUE/FALSE & WITH_BOOL_COMPAT define
2014-04-01 11:34:00 +11:00
static int panotestclip(Render *re, bool do_pano, float v[4])
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
{
fix for re->partx/re->party values being different for RE_Database_FromScene() then the main render loop.
2012-11-21 10:20:38 +00:00
/* part size (ensure we run RE_parts_clamp first) */
BLI_assert(re->partx == min_ii(re->r.tilex, re->rectx));
BLI_assert(re->party == min_ii(re->r.tiley, re->recty));
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
Code cleanup: remove TRUE/FALSE & WITH_BOOL_COMPAT define
2014-04-01 11:34:00 +11:00
if (do_pano == false) {
code cleanup: use TRUE/FALSE rather then 1/0 for better readability, also replace do prefix with do_ for bool vars.
2012-05-19 13:28:19 +00:00
return testclip(v);
}
fix for re->partx/re->party values being different for RE_Database_FromScene() then the main render loop.
2012-11-21 10:20:38 +00:00
else {
/* to be used for halos en infos */
float abs4;
short c = 0;
int xparts = (re->rectx + re->partx - 1) / re->partx;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
fix for re->partx/re->party values being different for RE_Database_FromScene() then the main render loop.
2012-11-21 10:20:38 +00:00
abs4= fabsf(v[3]);
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
fix for re->partx/re->party values being different for RE_Database_FromScene() then the main render loop.
2012-11-21 10:20:38 +00:00
if (v[2]< -abs4) c=16; /* this used to be " if (v[2]<0) ", see clippz() */
else if (v[2]> abs4) c+= 32;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
fix for re->partx/re->party values being different for RE_Database_FromScene() then the main render loop.
2012-11-21 10:20:38 +00:00
if ( v[1]>abs4) c+=4;
else if ( v[1]< -abs4) c+=8;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
fix for re->partx/re->party values being different for RE_Database_FromScene() then the main render loop.
2012-11-21 10:20:38 +00:00
abs4*= xparts;
if ( v[0]>abs4) c+=2;
else if ( v[0]< -abs4) c+=1;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
fix for re->partx/re->party values being different for RE_Database_FromScene() then the main render loop.
2012-11-21 10:20:38 +00:00
return c;
}
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
}
style cleanup: comment blocks
2012-03-09 18:28:30 +00:00
/**
* This adds the hcs coordinates to vertices. It iterates over all
* vertices, halos and faces. After the conversion, we clip in hcs.
*
* Elsewhere, all primites are converted to vertices.
* Called in
* - envmapping (envmap.c)
* - shadow buffering (shadbuf.c)
*/
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
define the size of matrix args for both rows/cols.
2012-12-11 14:29:01 +00:00
void project_renderdata(Render *re,
void (*projectfunc)(const float *, float mat[4][4], float *),
Code cleanup: remove TRUE/FALSE & WITH_BOOL_COMPAT define
2014-04-01 11:34:00 +11:00
bool do_pano, float xoffs, bool UNUSED(do_buckets))
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
{
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
ObjectRen *obr;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
HaloRen *har = NULL;
float zn, vec[3], hoco[4];
int a;
Particles ========= - Fix crash in particle transform with the particle system not editable. - Particle child distribution and caching is now multithreaded. - Child particles now have a separate Render Amount next to the existing Amount. The render amount particles are now only distributed and cached at render time, which should make editing with child particles faster. - Two new options for diffuse strand shading: - Surface Diffuse: computes the strand normal taking the normal at the surface into account. - Blending Distance: the distance in Blender units over which to blend in the normal at the surface. - Special strand rendering for more memory efficient and faster hair and grass. This is a work in progress, and has a number of known issues, don't report bugs to me for this feature yet. More info: http://www.blender.org/development/current-projects/changes-since-244/particles/
2007-12-04 13:57:28 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (do_pano) {
Recoded Panorama rendering. The old implementation was added quite hackish (talking about 10 yr ago). You also had to make a small image slice, which was extended Xparts in size. That also required to adjust the camera angle. Very clumsy. Now; when enabling the Panorama option, it will automatically apply the panorama effect on the vertically aligned tiles. You can just enable or disable the "Pano" button, to get a subtle lens effect like this: (without pano) http://www.blender.org/bf/rt.jpg (with pano) http://www.blender.org/bf/rt1.jpg For Panorama render, the minimum slice size has been hardcoded to be 8 pixels. The XParts button goes up to 512 to allow that. In practice, rendering 64 slices will already give very good images for a wide angle lens of 90 degrees, the curvature of straight lines then is equal to a circle of 256 points. Rendering a full 360 degree panorama you do by creating an extreme wide angle camera. The theory says camera-lens 5 should do 360 degrees, but for some reason my tests reveil it's 5.1... there's a rounding error somewhere, maybe related to the clipping plane start? Will look at that later. :) Also note that for each Xpart slice, the entire database needs to be rotated around camera to correct for panorama, on huge scenes that might give some overhead. Threaded render goes fine for Panorama too, but it can only render the vertically aligned parts in parallel. For the next panorama slice it has to wait for all threads of the current slice to be ready. On reading old files, I convert the settings to match as closely as possible the new situation. Since I cannot bump up the version #, the code detects for old panorama by checking for the image size. If image width is smaller than height, it assumes it's an old file (only if Panoroma option was set).
2006-02-27 12:39:36 +00:00
float panophi= xoffs;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
re->panosi= sin(panophi);
re->panoco= cos(panophi);
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (obr=re->objecttable.first; obr; obr=obr->next) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
/* calculate view coordinates (and zbuffer value) */
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (a=0; a<obr->tothalo; a++) {
if ((a & 255)==0) har= obr->bloha[a>>8];
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
else har++;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (do_pano) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
vec[0]= re->panoco*har->co[0] + re->panosi*har->co[2];
vec[1]= har->co[1];
vec[2]= -re->panosi*har->co[0] + re->panoco*har->co[2];
}
else {
replace VECCOPY -> copy_v3_v3, added copy_v*_v*_short too for typesafe copying, some parts of the code are copying float -> short normals without scaling. fix coming next.
2011-09-12 04:14:12 +00:00
copy_v3_v3(vec, har->co);
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
}
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
projectfunc(vec, re->winmat, hoco);
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
/* we clip halos less critical, but not for the Z */
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
hoco[0]*= 0.5f;
hoco[1]*= 0.5f;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if ( panotestclip(re, do_pano, hoco) ) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
har->miny= har->maxy= -10000; /* that way render clips it */
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (hoco[3]<0.0f) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
har->miny= har->maxy= -10000; /* render clips it */
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
}
style cleanup: changes to brace placement / newlines - for/while/if/switch
2012-04-28 06:31:57 +00:00
else { /* do the projection...*/
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
/* bring back hocos */
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
hoco[0]*= 2.0f;
hoco[1]*= 2.0f;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
zn= hoco[3];
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->xs= 0.5f*re->winx*(1.0f+hoco[0]/zn); /* the 0.5 negates the previous 2...*/
har->ys= 0.5f*re->winy*(1.0f+hoco[1]/zn);
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
/* this should be the zbuffer coordinate */
har->zs= 0x7FFFFF*(hoco[2]/zn);
/* taking this from the face clip functions? seems ok... */
har->zBufDist = 0x7FFFFFFF*(hoco[2]/zn);
vec[0]+= har->hasize;
projectfunc(vec, re->winmat, hoco);
vec[0]-= har->hasize;
zn= hoco[3];
floats were being promoted to doubles in quite a few cases (using gcc's -Wdouble-promotion), went over render module and use float constants, gives small but consistent speedup - approx 3%.
2011-08-20 17:39:13 +00:00
har->rad= fabsf(har->xs- 0.5f*re->winx*(1.0f+hoco[0]/zn));
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
/* this clip is not really OK, to prevent stars to become too large */
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (har->type & HA_ONLYSKY) {
if (har->rad>3.0f) har->rad= 3.0f;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
}
har->radsq= har->rad*har->rad;
har->miny= har->ys - har->rad/re->ycor;
har->maxy= har->ys + har->rad/re->ycor;
/* the Zd value is still not really correct for pano */
use fabsf when using floats.
2012-07-21 15:27:40 +00:00
vec[2] -= har->hasize; /* z negative, otherwise it's clipped */
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
projectfunc(vec, re->winmat, hoco);
use fabsf when using floats.
2012-07-21 15:27:40 +00:00
zn = hoco[3];
zn = fabsf((float)har->zs - 0x7FFFFF * (hoco[2] / zn));
har->zd = CLAMPIS(zn, 0, INT_MAX);
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
}
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
}
Render Baking ============= A new "Selected to Active" option in the Bake panel, to (typically) bake a high poly object onto a low poly object. Code based on patch #7339 by Frank Richter (Crystal Space developer), thanks!. Normal Mapping ============== Camera, World, Object and Tangent space is now supported for baking, and for material textures. The "NMap TS" setting is replaced with a dropdown of the four choices in the image texture buttons. http://www.blender.org/development/current-projects/changes-since-244/render-baking/
2007-11-28 22:21:12 +00:00
}
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
}
Initial revision
2002-10-12 11:37:38 +00:00
/* ------------------------------------------------------------------------- */
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
define the size of matrix args for both rows/cols.
2012-12-11 14:29:01 +00:00
ObjectInstanceRen *RE_addRenderInstance(Render *re, ObjectRen *obr, Object *ob, Object *par, int index, int psysindex, float mat[4][4], int lay)
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
{
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
ObjectInstanceRen *obi;
float mat3[3][3];
obi= MEM_callocN(sizeof(ObjectInstanceRen), "ObjectInstanceRen");
obi->obr= obr;
obi->ob= ob;
obi->par= par;
obi->index= index;
obi->psysindex= psysindex;
Bugfix: render instancing didn't work correct with layer ipo's, each instance should have it's own layer.
2008-03-11 12:29:59 +00:00
obi->lay= lay;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (mat) {
Math Lib * Convert all code to use new functions. * Branch maintainers may want to skip this commit, and run this conversion script instead, if they use a lot of math functions in new code: http://www.pasteall.org/9052/python
2009-11-10 20:43:45 +00:00
copy_m4_m4(obi->mat, mat);
copy_m3_m4(mat3, mat);
invert_m3_m3(obi->nmat, mat3);
transpose_m3(obi->nmat);
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obi->flag |= R_DUPLI_TRANSFORMED;
}
Subsurface scattering: Documentation on the settings, known limitations and implementation info can be found here: http://www.blender.org/development/current-projects/changes-since-243/subsurface-scattering/
2007-05-03 21:37:52 +00:00
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
BLI_addtail(&re->instancetable, obi);
Misc Render Features ==================== - "From Dupli" option for orco and uv texture coordinates. For dupliverts, duplifaces and dupli particles, this uses the orco and uv at the point on the parent surface. Can for example be used for texturing feathers and leafs. Note that uv only works for duplifaces and particles emitted from faces, these are not defined at vertices. - "Width Fade" option for strand render, to fade out along the width of the strand. Committing this so it can be tested, might be changed or removed even, if it doesn't give nice results.
2008-01-09 14:40:25 +00:00
return obi;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
}
void RE_makeRenderInstances(Render *re)
{
Dupli Bugfixes! It seems everytime I try to fix something here I break something else, but anyway, another try at getting this to work properly. Fixes for: - Getting dupliverts/faces derivedmesh with orco caused wrong results on meshes in linked dupligroups with proxy, because modifier stack was revaluated with wrong object matrix, now gets orco another way. - Fix render instances being hidden when original object was not added to the object render list. - Changed the way object instances find their original objects, now works the other way around, original objects look for their instances instead. There's probably issues still with recursive dupligroups..
2008-01-16 23:08:19 +00:00
ObjectInstanceRen *obi, *oldobi;
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
ListBase newlist;
int tot;
/* convert list of object instances to an array for index based lookup */
tot= BLI_countlist(&re->instancetable);
re->objectinstance= MEM_callocN(sizeof(ObjectInstanceRen)*tot, "ObjectInstance");
re->totinstance= tot;
newlist.first= newlist.last= NULL;
obi= re->objectinstance;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
for (oldobi=re->instancetable.first; oldobi; oldobi=oldobi->next) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
*obi= *oldobi;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (obi->obr) {
Render Instancing ================= Big commit, but little user visible changes. - Dupliverts and duplifaces are now rendered as instances, instead of storing all of the geometry for each dupli, now an instance is created with a matrix transform refering to the source object. This should allow us to render tree leaves more memory efficient. - Radiosity and to some degree raytracing of such objects is not really efficient still. For radiosity this is fundamentally hard to solve, but raytracing an octree could be created for each object, but the current octree code with it's fixed size doesn't allow this efficiently. - The regression tests survived, but with I expect that some bugs will pop up .. hopefully not too many :). Implementation Notes ==================== - Dupligroups and linked meshes are not rendered as instances yet, since they can in fact be different due to various reasons, instancing of these types of duplis that are the same can be added for them at a later point. - Each ObjectRen now stores it's own database, instead of there being one big databases of faces, verts, .. . Which objects that are actually rendered are defined by the list of ObjectRenInstances, which all refer to an ObjectRen. - Homogeneous coordinatess and clipping is now not stored in vertices anymore, but instead computed on the fly. This couldn't work for instances. That does mean some extra computation has to be done, but memory lookups can be slow too, and this saves some memory. Overall I didn't find a significant speed impact. - OSA rendering for solid and ztransp now is different. Instead of e.g. going 8 times over the databases times and rendering the z-buffer, it now goes over the database once and renders each polygon 8 times. That was necessary to keep instances efficient, and can also give some performance improvement without instances. - There was already instancing support in the yafray export code, now it uses Blender's render instances for export. - UV and color layer storage in the render was a bit messy before, now should be easier to understand. - convertblender.c was reorganized somewhat. Regular render, speedvector and baking now use a single function to create the database, previously there was code duplicated for it. - Some of these changes were done with future multithreading of scene and shadow buffer creation in mind, though especially for scene creation much work remains to be done to make it threadsafe, since it also involves a lot of code from blenkernel, and there is an ugly conflict with the way dupli groups work here .. though in the render code itself it's almost there.
2007-12-15 20:41:45 +00:00
obi->prev= obi->next= NULL;
BLI_addtail(&newlist, obi);
obi++;
}
else
re->totinstance--;
}
BLI_freelistN(&re->instancetable);
re->instancetable= newlist;
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
}
Fix T37675: blender internal viewport render not updating properly with volumes.
2013-12-18 15:41:28 +01:00
/* four functions to facilitate envmap rotation for raytrace */
void RE_instance_rotate_ray_start(ObjectInstanceRen *obi, Isect *is)
{
if (obi && (obi->flag & R_ENV_TRANSFORMED)) {
copy_v3_v3(is->origstart, is->start);
mul_m4_v3(obi->imat, is->start);
}
}
void RE_instance_rotate_ray_dir(ObjectInstanceRen *obi, Isect *is)
{
if (obi && (obi->flag & R_ENV_TRANSFORMED)) {
float end[3];
copy_v3_v3(is->origdir, is->dir);
add_v3_v3v3(end, is->origstart, is->dir);
mul_m4_v3(obi->imat, end);
sub_v3_v3v3(is->dir, end, is->start);
}
}
void RE_instance_rotate_ray(ObjectInstanceRen *obi, Isect *is)
{
RE_instance_rotate_ray_start(obi, is);
RE_instance_rotate_ray_dir(obi, is);
}
void RE_instance_rotate_ray_restore(ObjectInstanceRen *obi, Isect *is)
{
if (obi && (obi->flag & R_ENV_TRANSFORMED)) {
copy_v3_v3(is->start, is->origstart);
copy_v3_v3(is->dir, is->origdir);
}
}
define the size of matrix args for both rows/cols.
2012-12-11 14:29:01 +00:00
int clip_render_object(float boundbox[2][3], float bounds[4], float winmat[4][4])
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
{
Bounding box clipping in the render engine. Now bounding boxes are computed per object, and checked first before zbuffering objects. For strands, bounding boxes are computed per original face in the mesh. Overall the speed improvement from this is quite small (zbuffering is rarely the bottleneck), but it seems a sensible thing to do anyway.
2008-01-29 12:20:42 +00:00
float mat[4][4], vec[4];
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
int a, fl, flag = -1;
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
Math Lib * Convert all code to use new functions. * Branch maintainers may want to skip this commit, and run this conversion script instead, if they use a lot of math functions in new code: http://www.pasteall.org/9052/python
2009-11-10 20:43:45 +00:00
copy_m4_m4(mat, winmat);
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
for (a=0; a < 8; a++) {
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
vec[0]= (a & 1)? boundbox[0][0]: boundbox[1][0];
vec[1]= (a & 2)? boundbox[0][1]: boundbox[1][1];
vec[2]= (a & 4)? boundbox[0][2]: boundbox[1][2];
vec[3]= 1.0;
Math Lib * Convert all code to use new functions. * Branch maintainers may want to skip this commit, and run this conversion script instead, if they use a lot of math functions in new code: http://www.pasteall.org/9052/python
2009-11-10 20:43:45 +00:00
mul_m4_v4(mat, vec);
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
fl = 0;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
if (bounds) {
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
if (vec[0] < bounds[0] * vec[3]) fl |= 1;
else if (vec[0] > bounds[1] * vec[3]) fl |= 2;
Bugfix #24966 Hair render: using strand "Blender Unit" size didn't correctly clip for larger/wider strands. Now code clips strands based on the maximum width. Also found bad code for using clipping flags, which was mixed up, and probably caused hair strands to be missing in cases.
2010-12-26 17:47:17 +00:00
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
if (vec[1] > bounds[3] * vec[3]) fl |= 4;
else if (vec[1] < bounds[2] * vec[3]) fl |= 8;
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
}
else {
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
if (vec[0] < -vec[3]) fl |= 1;
else if (vec[0] > vec[3]) fl |= 2;
Bugfix #24966 Hair render: using strand "Blender Unit" size didn't correctly clip for larger/wider strands. Now code clips strands based on the maximum width. Also found bad code for using clipping flags, which was mixed up, and probably caused hair strands to be missing in cases.
2010-12-26 17:47:17 +00:00
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
if (vec[1] > vec[3]) fl |= 4;
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (vec[1] < -vec[3]) fl |= 8;
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
}
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
if (vec[2] < -vec[3]) fl |= 16;
else if (vec[2] > vec[3]) fl |= 32;
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
flag &= fl;
code cleanup: - define array sizes for functions that take vectors. - quiet some -Wshadow warnings. - some copy/paste error in readfile.c made it set the same particle recalc flag twice.
2012-10-20 08:02:18 +00:00
if (flag == 0) {
return 0;
}
Strands now mix together correctly with ZTransp. They now also store a list of samples per pixel, and then get shaded together with the ztransp samples. This comes with a slight speed hit, but mainly memory might be a concern. However, testing some peach scenes I haven't problems.
2008-01-28 16:54:52 +00:00
}
return flag;
}
Copy Permalink