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This repository has been archived on 2023-10-09.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
4119132580d8cf172e2b2ced86a7607d117b43c2
blender-archive/source/blender/render/intern/source/imagetexture.c

1796 lines
50 KiB
C
Raw Normal View History

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
/**
*
Fix syntax for ID keyword.
2010-03-21 01:14:04 +00:00
* $Id$
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
*
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 *****
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
*
* 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.
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
*
* 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.
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
*
* 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
* Contributors: 2004/2005/2006 Blender Foundation, full recode
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
*
* ***** END GPL/BL DUAL LICENSE BLOCK *****
*/
#include <stdio.h>
#include <string.h>
#include <fcntl.h>
#include <math.h>
Fix for bug #8750: depremul in the image texture code should check with an epsilon rather than 0.0f, could give random spots otherwise with e.g. calcalpha + negalpha.
2008-03-26 15:25:56 +00:00
#include <float.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
#ifndef WIN32
#include <unistd.h>
#else
#include <io.h>
#endif
#include "MEM_guardedalloc.h"
#include "IMB_imbuf_types.h"
#include "IMB_imbuf.h"
#include "DNA_image_types.h"
#include "DNA_scene_types.h"
#include "DNA_texture_types.h"
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"
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 "BLI_blenlib.h"
Fix for bug #5044: random crashes when rendering. This one is of my own making, removed too many locks. Thanks Malefico for reporting.
2006-09-30 21:29:47 +00:00
#include "BLI_threads.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
#include "BKE_utildefines.h"
#include "BKE_global.h"
#include "BKE_main.h"
#include "BKE_image.h"
#include "BKE_texture.h"
#include "BKE_library.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"
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 "texture.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
/* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */
/* defined in pipeline.c, is hardcopy of active dynamic allocated Render */
/* only to be used here in this file, it's for speed */
extern struct Render R;
/* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */
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
/* *********** IMAGEWRAPPING ****************** */
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
/* x and y have to be checked for image size */
static void ibuf_get_color(float *col, struct ImBuf *ibuf, int x, int y)
{
int ofs = y * ibuf->x + x;
if(ibuf->rect_float) {
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(ibuf->channels==4) {
float *fp= ibuf->rect_float + 4*ofs;
QUATCOPY(col, fp);
}
else if(ibuf->channels==3) {
float *fp= ibuf->rect_float + 3*ofs;
VECCOPY(col, fp);
col[3]= 1.0f;
}
else {
float *fp= ibuf->rect_float + ofs;
col[0]= col[1]= col[2]= col[3]= *fp;
}
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
}
else {
char *rect = (char *)( ibuf->rect+ ofs);
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
col[0] = ((float)rect[0])*(1.0f/255.0f);
col[1] = ((float)rect[1])*(1.0f/255.0f);
col[2] = ((float)rect[2])*(1.0f/255.0f);
col[3] = ((float)rect[3])*(1.0f/255.0f);
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
int imagewrap(Tex *tex, Image *ima, ImBuf *ibuf, float *texvec, TexResult *texres)
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 fx, fy, val1, val2, val3;
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
int x, y, retval;
Small fix for envmap rendering + threads... not sure if it fixes linux and windows for it... will be tested! Fix was that the envmap cube side images (6) were stored in tex->ima, which is shared data by other thread.
2005-04-25 20:36:17 +00:00
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
texres->tin= texres->ta= texres->tr= texres->tg= texres->tb= 0.0f;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* we need to set retval OK, otherwise texture code generates normals itself... */
retval= texres->nor?3:1;
/* quick tests */
if(ibuf==NULL && ima==NULL)
return retval;
if(ima) {
Bugfix #4557 Report noted the slow icon renders for menus, which indeed is an issue, especially when texture images need to be loaded (not to mention that will eat up loads of memory). Added a flag in scene to disable loading of images, makes it 50 times faster, at least :)
2006-07-02 18:26:20 +00:00
/* hack for icon render */
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(ima->ibufs.first==NULL && (R.r.scemode & R_NO_IMAGE_LOAD))
return retval;
ibuf= BKE_image_get_ibuf(ima, &tex->iuser);
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(ibuf==NULL || (ibuf->rect==NULL && ibuf->rect_float==NULL))
return retval;
/* setup mapping */
if(tex->imaflag & TEX_IMAROT) {
fy= texvec[0];
fx= texvec[1];
}
else {
fx= texvec[0];
fy= texvec[1];
}
if(tex->extend == TEX_CHECKER) {
int xs, ys;
xs= (int)floor(fx);
ys= (int)floor(fy);
fx-= xs;
fy-= ys;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if( (tex->flag & TEX_CHECKER_ODD)==0) {
if((xs+ys) & 1);else return retval;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if( (tex->flag & TEX_CHECKER_EVEN)==0) {
if((xs+ys) & 1) return retval;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* scale around center, (0.5, 0.5) */
if(tex->checkerdist<1.0) {
fx= (fx-0.5)/(1.0-tex->checkerdist) +0.5;
fy= (fy-0.5)/(1.0-tex->checkerdist) +0.5;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +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
bugfix [#24144] Small problem with texture (r32316) an extra pixel was being drawn on a tiled image texture.
2010-10-12 08:32:53 +00:00
x = (int)floorf(fx*ibuf->x);
y = (int)floorf(fy*ibuf->y);
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->extend == TEX_CLIPCUBE) {
if(x<0 || y<0 || x>=ibuf->x || y>=ibuf->y || texvec[2]<-1.0 || texvec[2]>1.0) {
return retval;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
}
else if( tex->extend==TEX_CLIP || tex->extend==TEX_CHECKER) {
if(x<0 || y<0 || x>=ibuf->x || y>=ibuf->y) {
return retval;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
}
else {
if(tex->extend==TEX_EXTEND) {
if(x>=ibuf->x) x = ibuf->x-1;
else if(x<0) x= 0;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
else {
x= x % ibuf->x;
if(x<0) x+= ibuf->x;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->extend==TEX_EXTEND) {
if(y>=ibuf->y) y = ibuf->y-1;
else if(y<0) y= 0;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
else {
y= y % ibuf->y;
if(y<0) y+= ibuf->y;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
}
/* warning, no return before setting back! */
if( (R.flag & R_SEC_FIELD) && (ibuf->flags & IB_fields) ) {
ibuf->rect+= (ibuf->x*ibuf->y);
}
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
ibuf_get_color(&texres->tr, ibuf, x, y);
if( (R.flag & R_SEC_FIELD) && (ibuf->flags & IB_fields) ) {
ibuf->rect-= (ibuf->x*ibuf->y);
}
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->imaflag & TEX_USEALPHA) {
if(tex->imaflag & TEX_CALCALPHA);
else texres->talpha= 1;
}
if(texres->nor) {
if(tex->imaflag & TEX_NORMALMAP) {
// qdn: normal from color
texres->nor[0] = 2.f*(texres->tr - 0.5f);
texres->nor[1] = 2.f*(0.5f - texres->tg);
texres->nor[2] = 2.f*(texres->tb - 0.5f);
}
else {
/* bump: take three samples */
val1= texres->tr+texres->tg+texres->tb;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(x<ibuf->x-1) {
float col[4];
ibuf_get_color(col, ibuf, x+1, y);
val2= (col[0]+col[1]+col[2]);
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
else val2= val1;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(y<ibuf->y-1) {
float col[4];
ibuf_get_color(col, ibuf, x, y+1);
val3= (col[0]+col[1]+col[2]);
}
else val3= val1;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* do not mix up x and y here! */
texres->nor[0]= (val1-val2);
texres->nor[1]= (val1-val3);
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(texres->talpha) texres->tin= texres->ta;
else if(tex->imaflag & TEX_CALCALPHA) {
texres->ta= texres->tin= MAX3(texres->tr, texres->tg, texres->tb);
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
else texres->ta= texres->tin= 1.0;
if(tex->flag & TEX_NEGALPHA) texres->ta= 1.0f-texres->ta;
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
Bugfix: de-premul for image texture didn't happen for non-osa case.
2007-12-04 19:06:00 +00:00
/* de-premul, this is being premulled in shade_input_do_shade() */
Fix for bug #8750: depremul in the image texture code should check with an epsilon rather than 0.0f, could give random spots otherwise with e.g. calcalpha + negalpha.
2008-03-26 15:25:56 +00:00
if(texres->ta!=1.0f && texres->ta>FLT_EPSILON) {
Bugfix: de-premul for image texture didn't happen for non-osa case.
2007-12-04 19:06:00 +00:00
fx= 1.0f/texres->ta;
texres->tr*= fx;
texres->tg*= fx;
texres->tb*= fx;
}
Fix [#22610] Alpha problem with textureswhen Brightness > 1 or Contrast < 1 * Enabled premultiplication for packed images * Added pack/unpack operator to image template * Moved brightness/contrast corrections to after de-premultiplication in image texture sampling
2010-06-17 07:20:12 +00:00
BRICONTRGB;
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
return retval;
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
}
static void clipx_rctf_swap(rctf *stack, short *count, float x1, float x2)
{
rctf *rf, *newrct;
short a;
a= *count;
rf= stack;
for(;a>0;a--) {
if(rf->xmin<x1) {
if(rf->xmax<x1) {
rf->xmin+= (x2-x1);
rf->xmax+= (x2-x1);
}
else {
if(rf->xmax>x2) rf->xmax= x2;
newrct= stack+ *count;
(*count)++;
newrct->xmax= x2;
newrct->xmin= rf->xmin+(x2-x1);
newrct->ymin= rf->ymin;
newrct->ymax= rf->ymax;
if(newrct->xmin==newrct->xmax) (*count)--;
rf->xmin= x1;
}
}
else if(rf->xmax>x2) {
if(rf->xmin>x2) {
rf->xmin-= (x2-x1);
rf->xmax-= (x2-x1);
}
else {
if(rf->xmin<x1) rf->xmin= x1;
newrct= stack+ *count;
(*count)++;
newrct->xmin= x1;
newrct->xmax= rf->xmax-(x2-x1);
newrct->ymin= rf->ymin;
newrct->ymax= rf->ymax;
if(newrct->xmin==newrct->xmax) (*count)--;
rf->xmax= x2;
}
}
rf++;
}
}
static void clipy_rctf_swap(rctf *stack, short *count, float y1, float y2)
{
rctf *rf, *newrct;
short a;
a= *count;
rf= stack;
for(;a>0;a--) {
if(rf->ymin<y1) {
if(rf->ymax<y1) {
rf->ymin+= (y2-y1);
rf->ymax+= (y2-y1);
}
else {
if(rf->ymax>y2) rf->ymax= y2;
newrct= stack+ *count;
(*count)++;
newrct->ymax= y2;
newrct->ymin= rf->ymin+(y2-y1);
newrct->xmin= rf->xmin;
newrct->xmax= rf->xmax;
if(newrct->ymin==newrct->ymax) (*count)--;
rf->ymin= y1;
}
}
else if(rf->ymax>y2) {
if(rf->ymin>y2) {
rf->ymin-= (y2-y1);
rf->ymax-= (y2-y1);
}
else {
if(rf->ymin<y1) rf->ymin= y1;
newrct= stack+ *count;
(*count)++;
newrct->ymin= y1;
newrct->ymax= rf->ymax-(y2-y1);
newrct->xmin= rf->xmin;
newrct->xmax= rf->xmax;
if(newrct->ymin==newrct->ymax) (*count)--;
rf->ymax= y2;
}
}
rf++;
}
}
static float square_rctf(rctf *rf)
{
float x, y;
x= rf->xmax- rf->xmin;
y= rf->ymax- rf->ymin;
return (x*y);
}
static float clipx_rctf(rctf *rf, float x1, float x2)
{
float size;
size= rf->xmax - rf->xmin;
if(rf->xmin<x1) {
rf->xmin= x1;
}
if(rf->xmax>x2) {
rf->xmax= x2;
}
if(rf->xmin > rf->xmax) {
rf->xmin = rf->xmax;
return 0.0;
}
else if(size!=0.0) {
return (rf->xmax - rf->xmin)/size;
}
return 1.0;
}
static float clipy_rctf(rctf *rf, float y1, float y2)
{
float size;
size= rf->ymax - rf->ymin;
if(rf->ymin<y1) {
rf->ymin= y1;
}
if(rf->ymax>y2) {
rf->ymax= y2;
}
if(rf->ymin > rf->ymax) {
rf->ymin = rf->ymax;
return 0.0;
}
else if(size!=0.0) {
return (rf->ymax - rf->ymin)/size;
}
return 1.0;
}
static void boxsampleclip(struct ImBuf *ibuf, rctf *rf, TexResult *texres)
{
/* sample box, is clipped already, and minx etc. have been set at ibuf size.
spaces -> tabs, (4 spaces == 1 tab, only for white space preceding text)
2010-03-22 09:30:00 +00:00
Enlarge with antialiased edges of the pixels */
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
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
float muly, mulx, div, col[4];
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
int x, y, startx, endx, starty, endy;
startx= (int)floor(rf->xmin);
endx= (int)floor(rf->xmax);
starty= (int)floor(rf->ymin);
endy= (int)floor(rf->ymax);
if(startx < 0) startx= 0;
if(starty < 0) starty= 0;
if(endx>=ibuf->x) endx= ibuf->x-1;
if(endy>=ibuf->y) endy= ibuf->y-1;
if(starty==endy && startx==endx) {
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
ibuf_get_color(&texres->tr, ibuf, startx, starty);
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
}
else {
div= texres->tr= texres->tg= texres->tb= texres->ta= 0.0;
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
for(y=starty; y<=endy; y++) {
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
muly= 1.0;
if(starty==endy);
else {
if(y==starty) muly= 1.0f-(rf->ymin - y);
if(y==endy) muly= (rf->ymax - y);
}
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +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
if(startx==endx) {
mulx= muly;
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
ibuf_get_color(col, ibuf, startx, y);
texres->ta+= mulx*col[3];
texres->tr+= mulx*col[0];
texres->tg+= mulx*col[1];
texres->tb+= mulx*col[2];
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
div+= mulx;
}
else {
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
for(x=startx; x<=endx; x++) {
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
mulx= muly;
if(x==startx) mulx*= 1.0f-(rf->xmin - x);
if(x==endx) mulx*= (rf->xmax - x);
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
ibuf_get_color(col, ibuf, x, y);
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
if(mulx==1.0) {
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
texres->ta+= col[3];
texres->tr+= col[0];
texres->tg+= col[1];
texres->tb+= col[2];
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
div+= 1.0;
}
else {
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
texres->ta+= mulx*col[3];
texres->tr+= mulx*col[0];
texres->tg+= mulx*col[1];
texres->tb+= mulx*col[2];
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
div+= mulx;
}
}
}
}
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +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
if(div!=0.0) {
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
div= 1.0f/div;
texres->tb*= div;
texres->tg*= div;
texres->tr*= div;
texres->ta*= div;
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
}
else {
Orange: more float buffer support; - Image textures use float colors now, when present. Works for mipmap too, and for AO skycolor (probes) - Backbuffer option uses float buffers too. Note that rendering OSA will resample the backbuffer, filtering it... will need to be solved with the new composit stage - LMB sampling in image window now shows float color too + bugfix in imbuf, filtering for float buffers had an error.
2006-01-11 18:39:19 +00:00
texres->tr= texres->tg= texres->tb= texres->ta= 0.0f;
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
}
}
}
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
static void boxsample(ImBuf *ibuf, float minx, float miny, float maxx, float maxy, TexResult *texres, int imaprepeat, int imapextend)
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
{
/* Sample box, performs clip. minx etc are in range 0.0 - 1.0 .
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
* Enlarge with antialiased edges of pixels.
* If variable 'imaprepeat' has been set, the
* clipped-away parts are sampled as well.
*/
Bugfix #3463 Very ancient bug in boxfilter for image texture popped up with new code for correct bumpmapping. Caused texture type "Extend" not to work when using for bump.
2005-11-26 17:56:39 +00:00
/* note: actually minx etc isnt in the proper range... this due to filter size and offset vectors for bump */
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
TexResult texr;
rctf *rf, stack[8];
float opp, tot, alphaclip= 1.0;
short count=1;
rf= stack;
rf->xmin= minx*(ibuf->x);
rf->xmax= maxx*(ibuf->x);
rf->ymin= miny*(ibuf->y);
rf->ymax= maxy*(ibuf->y);
#3598 Uninitialized 'alpha' flag in Image texture could cause render errors. Didn't show in OSX though... (stupid system here inits stack to zero). But, Ken Hughes found this, so all credits are for him!
2005-12-19 11:30:52 +00:00
texr.talpha= texres->talpha; /* is read by boxsample_clip */
Bugfix #3463 Very ancient bug in boxfilter for image texture popped up with new code for correct bumpmapping. Caused texture type "Extend" not to work when using for bump.
2005-11-26 17:56:39 +00:00
if(imapextend) {
CLAMP(rf->xmin, 0.0f, ibuf->x-1);
CLAMP(rf->xmax, 0.0f, ibuf->x-1);
}
Bugfix #5962 New Image texture option for "Repeat Mirror" didnt work when the texture "Map input" had a translation.
2007-02-09 10:35:04 +00:00
else if(imaprepeat)
clipx_rctf_swap(stack, &count, 0.0, (float)(ibuf->x));
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
else {
alphaclip= clipx_rctf(rf, 0.0, (float)(ibuf->x));
if(alphaclip<=0.0) {
texres->tr= texres->tb= texres->tg= texres->ta= 0.0;
return;
}
}
Bugfix #3463 Very ancient bug in boxfilter for image texture popped up with new code for correct bumpmapping. Caused texture type "Extend" not to work when using for bump.
2005-11-26 17:56:39 +00:00
if(imapextend) {
CLAMP(rf->ymin, 0.0f, ibuf->y-1);
CLAMP(rf->ymax, 0.0f, ibuf->y-1);
}
Bugfix #5962 New Image texture option for "Repeat Mirror" didnt work when the texture "Map input" had a translation.
2007-02-09 10:35:04 +00:00
else if(imaprepeat)
clipy_rctf_swap(stack, &count, 0.0, (float)(ibuf->y));
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
else {
alphaclip*= clipy_rctf(rf, 0.0, (float)(ibuf->y));
if(alphaclip<=0.0) {
texres->tr= texres->tb= texres->tg= texres->ta= 0.0;
return;
}
}
if(count>1) {
tot= texres->tr= texres->tb= texres->tg= texres->ta= 0.0;
while(count--) {
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
boxsampleclip(ibuf, rf, &texr);
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
opp= square_rctf(rf);
tot+= opp;
texres->tr+= opp*texr.tr;
texres->tg+= opp*texr.tg;
texres->tb+= opp*texr.tb;
if(texres->talpha) texres->ta+= opp*texr.ta;
rf++;
}
if(tot!= 0.0) {
texres->tr/= tot;
texres->tg/= tot;
texres->tb/= tot;
if(texres->talpha) texres->ta/= tot;
}
}
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
else
boxsampleclip(ibuf, rf, texres);
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
if(texres->talpha==0) texres->ta= 1.0;
if(alphaclip!=1.0) {
Bugfix #14033 Option for Image render "Clip" gave wrong alphas on the edge of image.
2008-09-11 16:50:34 +00:00
/* premul it all */
texres->tr*= alphaclip;
texres->tg*= alphaclip;
texres->tb*= alphaclip;
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
texres->ta*= alphaclip;
}
}
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
//-----------------------------------------------------------------------------------------------------------------
// from here, some functions only used for the new filtering
Fix [#22418] Displace Node also makes entire image fuzzy Changed displace node sampling to use EWA filtering, and removed old hacks for calculating derivatives - I think it should be generated correctly now.
2010-05-26 08:06:51 +00:00
// anisotropic filters, data struct used instead of long line of (possibly unused) func args
typedef struct afdata_t {
float dxt[2], dyt[2];
int intpol, extflag;
// feline only
float majrad, minrad, theta;
int iProbes;
float dusc, dvsc;
} afdata_t;
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
// this only used here to make it easier to pass extend flags as single int
enum {TXC_XMIR=1, TXC_YMIR, TXC_REPT, TXC_EXTD};
// similar to ibuf_get_color() but clips/wraps coords according to repeat/extend flags
// returns true if out of range in clipmode
static int ibuf_get_color_clip(float *col, ImBuf *ibuf, int x, int y, int extflag)
{
int clip = 0;
switch (extflag) {
case TXC_XMIR: // y rep
x %= 2*ibuf->x;
x += x < 0 ? 2*ibuf->x : 0;
x = x >= ibuf->x ? 2*ibuf->x - x - 1 : x;
y %= ibuf->y;
y += y < 0 ? ibuf->y : 0;
break;
case TXC_YMIR: // x rep
x %= ibuf->x;
x += x < 0 ? ibuf->x : 0;
y %= 2*ibuf->y;
y += y < 0 ? 2*ibuf->y : 0;
y = y >= ibuf->y ? 2*ibuf->y - y - 1 : y;
break;
case TXC_EXTD:
x = (x < 0) ? 0 : ((x >= ibuf->x) ? (ibuf->x - 1) : x);
y = (y < 0) ? 0 : ((y >= ibuf->y) ? (ibuf->y - 1) : y);
break;
case TXC_REPT:
x %= ibuf->x;
x += (x < 0) ? ibuf->x : 0;
y %= ibuf->y;
y += (y < 0) ? ibuf->y : 0;
break;
default: { // as extend, if clipped, set alpha to 0.0
if (x < 0) { x = 0; } // TXF alpha: clip = 1; }
if (x >= ibuf->x) { x = ibuf->x - 1; } // TXF alpha: clip = 1; }
if (y < 0) { y = 0; } // TXF alpha: clip = 1; }
if (y >= ibuf->y) { y = ibuf->y - 1; } // TXF alpha: clip = 1; }
}
}
if (ibuf->rect_float) {
const float* fp = ibuf->rect_float + (x + y*ibuf->x)*ibuf->channels;
if (ibuf->channels == 1)
col[0] = col[1] = col[2] = col[3] = *fp;
else {
col[0] = fp[0];
col[1] = fp[1];
col[2] = fp[2];
col[3] = clip ? 0.f : (ibuf->channels == 4 ? fp[3] : 1.f);
}
}
else {
char* rect = (char*)(ibuf->rect + x + y*ibuf->x);
col[0] = rect[0]*(1.f/255.f);
col[1] = rect[1]*(1.f/255.f);
col[2] = rect[2]*(1.f/255.f);
col[3] = clip ? 0.f : rect[3]*(1.f/255.f);
}
return clip;
}
// as above + bilerp
static int ibuf_get_color_clip_bilerp(float *col, ImBuf *ibuf, float u, float v, int intpol, int extflag)
{
if (intpol) {
float c00[4], c01[4], c10[4], c11[4];
const float ufl = floorf(u -= 0.5f), vfl = floorf(v -= 0.5f);
const float uf = u - ufl, vf = v - vfl;
const float w00=(1.f-uf)*(1.f-vf), w10=uf*(1.f-vf), w01=(1.f-uf)*vf, w11=uf*vf;
const int x1 = (int)ufl, y1 = (int)vfl, x2 = x1 + 1, y2 = y1 + 1;
int clip = ibuf_get_color_clip(c00, ibuf, x1, y1, extflag);
clip |= ibuf_get_color_clip(c10, ibuf, x2, y1, extflag);
clip |= ibuf_get_color_clip(c01, ibuf, x1, y2, extflag);
clip |= ibuf_get_color_clip(c11, ibuf, x2, y2, extflag);
col[0] = w00*c00[0] + w10*c10[0] + w01*c01[0] + w11*c11[0];
col[1] = w00*c00[1] + w10*c10[1] + w01*c01[1] + w11*c11[1];
col[2] = w00*c00[2] + w10*c10[2] + w01*c01[2] + w11*c11[2];
col[3] = clip ? 0.f : w00*c00[3] + w10*c10[3] + w01*c01[3] + w11*c11[3];
return clip;
}
return ibuf_get_color_clip(col, ibuf, (int)u, (int)v, extflag);
}
static void area_sample(TexResult* texr, ImBuf* ibuf, float fx, float fy, afdata_t* AFD)
{
int xs, ys, clip = 0;
float tc[4], xsd, ysd, cw = 0.f;
const float ux = ibuf->x*AFD->dxt[0], uy = ibuf->y*AFD->dxt[1];
const float vx = ibuf->x*AFD->dyt[0], vy = ibuf->y*AFD->dyt[1];
int xsam = (int)(0.5f*sqrtf(ux*ux + uy*uy) + 0.5f);
int ysam = (int)(0.5f*sqrtf(vx*vx + vy*vy) + 0.5f);
const int minsam = AFD->intpol ? 2 : 4;
Fix very long render time when using a large filter size with the new image sample code. I've added a dumb clamping now of dxt/dyt, probably could use a smarter solution but it's not likely to be noticeable.
2010-02-15 08:29:01 +00:00
xsam = CLAMPIS(xsam, minsam, ibuf->x*2);
ysam = CLAMPIS(ysam, minsam, ibuf->y*2);
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
xsd = 1.f / xsam;
ysd = 1.f / ysam;
texr->tr = texr->tg = texr->tb = texr->ta = 0.f;
for (ys=0; ys<ysam; ++ys) {
for (xs=0; xs<xsam; ++xs) {
const float su = (xs + ((ys & 1) + 0.5f)*0.5f)*xsd - 0.5f;
const float sv = (ys + ((xs & 1) + 0.5f)*0.5f)*ysd - 0.5f;
const float pu = fx + su*AFD->dxt[0] + sv*AFD->dyt[0];
const float pv = fy + su*AFD->dxt[1] + sv*AFD->dyt[1];
const int out = ibuf_get_color_clip_bilerp(tc, ibuf, pu*ibuf->x, pv*ibuf->y, AFD->intpol, AFD->extflag);
clip |= out;
cw += out ? 0.f : 1.f;
texr->tr += tc[0];
texr->tg += tc[1];
texr->tb += tc[2];
texr->ta += texr->talpha ? tc[3] : 0.f;
}
}
xsd *= ysd;
texr->tr *= xsd;
texr->tg *= xsd;
texr->tb *= xsd;
// clipping can be ignored if alpha used, texr->ta already includes filtered edge
texr->ta = texr->talpha ? texr->ta*xsd : (clip ? cw*xsd : 1.f);
}
// table of (exp(ar) - exp(a)) / (1 - exp(a)) for r in range [0, 1] and a = -2
// used instead of actual gaussian, otherwise at high texture magnifications circular artifacts are visible
#define EWA_MAXIDX 255
static float EWA_WTS[EWA_MAXIDX + 1] =
{ 1.f, 0.990965f, 0.982f, 0.973105f, 0.96428f, 0.955524f, 0.946836f, 0.938216f, 0.929664f,
0.921178f, 0.912759f, 0.904405f, 0.896117f, 0.887893f, 0.879734f, 0.871638f, 0.863605f,
0.855636f, 0.847728f, 0.839883f, 0.832098f, 0.824375f, 0.816712f, 0.809108f, 0.801564f,
0.794079f, 0.786653f, 0.779284f, 0.771974f, 0.76472f, 0.757523f, 0.750382f, 0.743297f,
0.736267f, 0.729292f, 0.722372f, 0.715505f, 0.708693f, 0.701933f, 0.695227f, 0.688572f,
0.68197f, 0.67542f, 0.66892f, 0.662471f, 0.656073f, 0.649725f, 0.643426f, 0.637176f,
0.630976f, 0.624824f, 0.618719f, 0.612663f, 0.606654f, 0.600691f, 0.594776f, 0.588906f,
0.583083f, 0.577305f, 0.571572f, 0.565883f, 0.56024f, 0.55464f, 0.549084f, 0.543572f,
0.538102f, 0.532676f, 0.527291f, 0.521949f, 0.516649f, 0.511389f, 0.506171f, 0.500994f,
0.495857f, 0.490761f, 0.485704f, 0.480687f, 0.475709f, 0.470769f, 0.465869f, 0.461006f,
0.456182f, 0.451395f, 0.446646f, 0.441934f, 0.437258f, 0.432619f, 0.428017f, 0.42345f,
0.418919f, 0.414424f, 0.409963f, 0.405538f, 0.401147f, 0.39679f, 0.392467f, 0.388178f,
0.383923f, 0.379701f, 0.375511f, 0.371355f, 0.367231f, 0.363139f, 0.359079f, 0.355051f,
0.351055f, 0.347089f, 0.343155f, 0.339251f, 0.335378f, 0.331535f, 0.327722f, 0.323939f,
0.320186f, 0.316461f, 0.312766f, 0.3091f, 0.305462f, 0.301853f, 0.298272f, 0.294719f,
0.291194f, 0.287696f, 0.284226f, 0.280782f, 0.277366f, 0.273976f, 0.270613f, 0.267276f,
0.263965f, 0.26068f, 0.257421f, 0.254187f, 0.250979f, 0.247795f, 0.244636f, 0.241502f,
0.238393f, 0.235308f, 0.232246f, 0.229209f, 0.226196f, 0.223206f, 0.220239f, 0.217296f,
0.214375f, 0.211478f, 0.208603f, 0.20575f, 0.20292f, 0.200112f, 0.197326f, 0.194562f,
0.191819f, 0.189097f, 0.186397f, 0.183718f, 0.18106f, 0.178423f, 0.175806f, 0.17321f,
0.170634f, 0.168078f, 0.165542f, 0.163026f, 0.16053f, 0.158053f, 0.155595f, 0.153157f,
0.150738f, 0.148337f, 0.145955f, 0.143592f, 0.141248f, 0.138921f, 0.136613f, 0.134323f,
0.132051f, 0.129797f, 0.12756f, 0.125341f, 0.123139f, 0.120954f, 0.118786f, 0.116635f,
0.114501f, 0.112384f, 0.110283f, 0.108199f, 0.106131f, 0.104079f, 0.102043f, 0.100023f,
0.0980186f, 0.09603f, 0.094057f, 0.0920994f, 0.0901571f, 0.08823f, 0.0863179f, 0.0844208f,
0.0825384f, 0.0806708f, 0.0788178f, 0.0769792f, 0.0751551f, 0.0733451f, 0.0715493f, 0.0697676f,
0.0679997f, 0.0662457f, 0.0645054f, 0.0627786f, 0.0610654f, 0.0593655f, 0.0576789f, 0.0560055f,
0.0543452f, 0.0526979f, 0.0510634f, 0.0494416f, 0.0478326f, 0.0462361f, 0.0446521f, 0.0430805f,
0.0415211f, 0.039974f, 0.0384389f, 0.0369158f, 0.0354046f, 0.0339052f, 0.0324175f, 0.0309415f,
0.029477f, 0.0280239f, 0.0265822f, 0.0251517f, 0.0237324f, 0.0223242f, 0.020927f, 0.0195408f,
0.0181653f, 0.0168006f, 0.0154466f, 0.0141031f, 0.0127701f, 0.0114476f, 0.0101354f, 0.00883339f,
0.00754159f, 0.00625989f, 0.00498819f, 0.00372644f, 0.00247454f, 0.00123242f, 0.f
};
// test if a float value is 'nan'
// there is a C99 function for this: isnan(), but blender seems to use C90 (according to gcc warns),
// and may not be supported by other compilers either
#ifndef ISNAN
#define ISNAN(x) ((x) != (x))
#endif
//static int ISNAN(float x) { return (x != x); }
static void radangle2imp(float a2, float b2, float th, float* A, float* B, float* C, float* F)
{
float ct2 = cosf(th);
const float st2 = 1.f - ct2*ct2; // <- sin(th)^2
ct2 *= ct2;
*A = a2*st2 + b2*ct2;
*B = (b2 - a2)*sinf(2.f*th);
*C = a2*ct2 + b2*st2;
*F = a2*b2;
}
// all tests here are done to make sure possible overflows are hopefully minimized
static void imp2radangle(float A, float B, float C, float F, float* a, float* b, float* th, float* ecc)
{
if (F <= 1e-5f) { // use arbitrary major radius, zero minor, infinite eccentricity
*a = sqrtf(A > C ? A : C);
*b = 0.f;
*ecc = 1e10f;
*th = 0.5f*(atan2f(B, A - C) + (float)M_PI);
}
else {
const float AmC = A - C, ApC = A + C, F2 = F*2.f;
const float r = sqrtf(AmC*AmC + B*B);
float d = ApC - r;
*a = (d <= 0.f) ? sqrtf(A > C ? A : C) : sqrtf(F2 / d);
d = ApC + r;
if (d <= 0.f) {
*b = 0.f;
*ecc = 1e10f;
}
else {
*b = sqrtf(F2 / d);
*ecc = *a / *b;
}
// incr theta by 0.5*pi (angle of major axis)
*th = 0.5f*(atan2f(B, AmC) + (float)M_PI);
}
}
static void ewa_eval(TexResult* texr, ImBuf* ibuf, float fx, float fy, afdata_t* AFD)
{
// scaling dxt/dyt by full resolution can cause overflow because of huge A/B/C and esp. F values,
// scaling by aspect ratio alone does the opposite, so try something inbetween instead...
const float ff2 = ibuf->x, ff = sqrtf(ff2), q = ibuf->y / ff;
const float Ux = AFD->dxt[0]*ff, Vx = AFD->dxt[1]*q, Uy = AFD->dyt[0]*ff, Vy = AFD->dyt[1]*q;
float A = Vx*Vx + Vy*Vy;
float B = -2.f*(Ux*Vx + Uy*Vy);
float C = Ux*Ux + Uy*Uy;
float F = A*C - B*B*0.25f;
float a, b, th, ecc, a2, b2, ue, ve, U0, V0, DDQ, U, ac1, ac2, BU, d; // TXF alpha: cw = 0.f;
int u, v, u1, u2, v1, v2; // TXF alpha: clip = 0;
// The so-called 'high' quality ewa method simply adds a constant of 1 to both A & C,
// so the ellipse always covers at least some texels. But since the filter is now always larger,
// it also means that everywhere else it's also more blurry then ideally should be the case.
// So instead here the ellipse radii are modified instead whenever either is too low.
// Use a different radius based on interpolation switch, just enough to anti-alias when interpolation is off,
// and slightly larger to make result a bit smoother than bilinear interpolation when interpolation is on
// (minimum values: const float rmin = intpol ? 1.f : 0.5f;)
const float rmin = (AFD->intpol ? 1.5625f : 0.765625f)/ff2;
imp2radangle(A, B, C, F, &a, &b, &th, &ecc);
if ((b2 = b*b) < rmin) {
if ((a2 = a*a) < rmin) {
B = 0.f;
A = C = rmin;
F = A*C;
}
else {
b2 = rmin;
radangle2imp(a2, b2, th, &A, &B, &C, &F);
}
}
ue = ff*sqrtf(C);
ve = ff*sqrtf(A);
d = (float)(EWA_MAXIDX + 1) / (F*ff2);
A *= d;
B *= d;
C *= d;
U0 = fx*ibuf->x;
V0 = fy*ibuf->y;
u1 = (int)(floorf(U0 - ue));
u2 = (int)(ceilf(U0 + ue));
v1 = (int)(floorf(V0 - ve));
v2 = (int)(ceilf(V0 + ve));
U0 -= 0.5f;
V0 -= 0.5f;
DDQ = 2.f*A;
U = u1 - U0;
ac1 = A*(2.f*U + 1.f);
ac2 = A*U*U;
BU = B*U;
d = texr->tr = texr->tb = texr->tg = texr->ta = 0.f;
for (v=v1; v<=v2; ++v) {
const float V = v - V0;
float DQ = ac1 + B*V;
float Q = (C*V + BU)*V + ac2;
for (u=u1; u<=u2; ++u) {
if (Q < (float)(EWA_MAXIDX + 1)) {
float tc[4];
const float wt = EWA_WTS[(Q < 0.f) ? 0 : (unsigned int)Q];
/*const int out =*/ ibuf_get_color_clip(tc, ibuf, u, v, AFD->extflag);
// TXF alpha: clip |= out;
// TXF alpha: cw += out ? 0.f : wt;
texr->tr += tc[0]*wt;
texr->tg += tc[1]*wt;
texr->tb += tc[2]*wt;
texr->ta += texr->talpha ? tc[3]*wt : 0.f;
d += wt;
}
Q += DQ;
DQ += DDQ;
}
}
// d should hopefully never be zero anymore
d = 1.f/d;
texr->tr *= d;
texr->tg *= d;
texr->tb *= d;
// clipping can be ignored if alpha used, texr->ta already includes filtered edge
texr->ta = texr->talpha ? texr->ta*d : 1.f; // TXF alpha (clip ? cw*d : 1.f);
}
static void feline_eval(TexResult* texr, ImBuf* ibuf, float fx, float fy, afdata_t* AFD)
{
const int maxn = AFD->iProbes - 1;
const float ll = ((AFD->majrad == AFD->minrad) ? 2.f*AFD->majrad : 2.f*(AFD->majrad - AFD->minrad)) / (maxn ? (float)maxn : 1.f);
float du = maxn ? cosf(AFD->theta)*ll : 0.f;
float dv = maxn ? sinf(AFD->theta)*ll : 0.f;
//const float D = -0.5f*(du*du + dv*dv) / (AFD->majrad*AFD->majrad);
const float D = (EWA_MAXIDX + 1)*0.25f*(du*du + dv*dv) / (AFD->majrad*AFD->majrad);
float d; // TXF alpha: cw = 0.f;
int n; // TXF alpha: clip = 0;
// have to use same scaling for du/dv here as for Ux/Vx/Uy/Vy (*after* D calc.)
du *= AFD->dusc;
dv *= AFD->dvsc;
d = texr->tr = texr->tb = texr->tg = texr->ta = 0.f;
for (n=-maxn; n<=maxn; n+=2) {
float tc[4];
const float hn = n*0.5f;
const float u = fx + hn*du, v = fy + hn*dv;
//const float wt = expf(n*n*D);
// can use ewa table here too
const float wt = EWA_WTS[(int)(n*n*D)];
/*const int out =*/ ibuf_get_color_clip_bilerp(tc, ibuf, ibuf->x*u, ibuf->y*v, AFD->intpol, AFD->extflag);
// TXF alpha: clip |= out;
// TXF alpha: cw += out ? 0.f : wt;
texr->tr += tc[0]*wt;
texr->tg += tc[1]*wt;
texr->tb += tc[2]*wt;
texr->ta += texr->talpha ? tc[3]*wt : 0.f;
d += wt;
}
d = 1.f/d;
texr->tr *= d;
texr->tg *= d;
texr->tb *= d;
// clipping can be ignored if alpha used, texr->ta already includes filtered edge
texr->ta = texr->talpha ? texr->ta*d : 1.f; // TXF alpha: (clip ? cw*d : 1.f);
}
#undef EWA_MAXIDX
static void alpha_clip_aniso(ImBuf *ibuf, float minx, float miny, float maxx, float maxy, int extflag, TexResult *texres)
{
float alphaclip;
rctf rf;
// TXF apha: we're doing the same alphaclip here as boxsample, but i'm doubting
// if this is actually correct for the all the filtering algorithms ..
if(!(extflag == TXC_REPT || extflag == TXC_EXTD)) {
rf.xmin= minx*(ibuf->x);
rf.xmax= maxx*(ibuf->x);
rf.ymin= miny*(ibuf->y);
rf.ymax= maxy*(ibuf->y);
alphaclip = clipx_rctf(&rf, 0.0, (float)(ibuf->x));
alphaclip*= clipy_rctf(&rf, 0.0, (float)(ibuf->y));
alphaclip= MAX2(alphaclip, 0.0f);
if(alphaclip!=1.0) {
/* premul it all */
texres->tr*= alphaclip;
texres->tg*= alphaclip;
texres->tb*= alphaclip;
texres->ta*= alphaclip;
}
}
}
Bugfix #22040 Old bug report: Image Editor, Painting: crash when texture was visible in Material or Texture preview. Was caused by paint code freeing mipmaps. Now replaced with a mipmap tag (to be done again), and a new mipmap function that doesn't re-allocate.
2010-12-14 18:02:41 +00:00
static void image_mipmap_test(Tex *tex, ImBuf *ibuf)
{
if (tex->imaflag & TEX_MIPMAP) {
if ((ibuf->flags & IB_fields) == 0) {
if (ibuf->mipmap[0] && (ibuf->userflags & IB_MIPMAP_INVALID)) {
BLI_lock_thread(LOCK_IMAGE);
if (ibuf->userflags & IB_MIPMAP_INVALID) {
IMB_remakemipmap(ibuf, tex->imaflag & TEX_GAUSS_MIP);
ibuf->userflags &= ~IB_MIPMAP_INVALID;
}
BLI_unlock_thread(LOCK_IMAGE);
}
if (ibuf->mipmap[0] == NULL) {
BLI_lock_thread(LOCK_IMAGE);
if (ibuf->mipmap[0] == NULL)
IMB_makemipmap(ibuf, tex->imaflag & TEX_GAUSS_MIP);
BLI_unlock_thread(LOCK_IMAGE);
}
}
}
}
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
static int imagewraposa_aniso(Tex *tex, Image *ima, ImBuf *ibuf, float *texvec, float *dxt, float *dyt, TexResult *texres)
{
TexResult texr;
float fx, fy, minx, maxx, miny, maxy;
float maxd, val1, val2, val3;
int curmap, retval, intpol, extflag = 0;
afdata_t AFD;
void (*filterfunc)(TexResult*, ImBuf*, float, float, afdata_t*);
switch (tex->texfilter) {
case TXF_EWA:
filterfunc = ewa_eval;
break;
case TXF_FELINE:
filterfunc = feline_eval;
break;
case TXF_AREA:
default:
filterfunc = area_sample;
}
texres->tin = texres->ta = texres->tr = texres->tg = texres->tb = 0.f;
// we need to set retval OK, otherwise texture code generates normals itself...
retval = texres->nor ? 3 : 1;
// quick tests
if (ibuf==NULL && ima==NULL) return retval;
if (ima) { // hack for icon render
if ((ima->ibufs.first == NULL) && (R.r.scemode & R_NO_IMAGE_LOAD)) return retval;
ibuf = BKE_image_get_ibuf(ima, &tex->iuser);
}
if ((ibuf == NULL) || ((ibuf->rect == NULL) && (ibuf->rect_float == NULL))) return retval;
Bugfix #22040 Old bug report: Image Editor, Painting: crash when texture was visible in Material or Texture preview. Was caused by paint code freeing mipmaps. Now replaced with a mipmap tag (to be done again), and a new mipmap function that doesn't re-allocate.
2010-12-14 18:02:41 +00:00
/* mipmap test */
image_mipmap_test(tex, ibuf);
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
if ((tex->imaflag & TEX_USEALPHA) && ((tex->imaflag & TEX_CALCALPHA) == 0)) texres->talpha = 1;
texr.talpha = texres->talpha;
if (tex->imaflag & TEX_IMAROT) {
fy = texvec[0];
fx = texvec[1];
}
else {
fx = texvec[0];
fy = texvec[1];
}
if (ibuf->flags & IB_fields) {
if (R.r.mode & R_FIELDS) { /* field render */
if (R.flag & R_SEC_FIELD) { /* correction for 2nd field */
/* fac1= 0.5/( (float)ibuf->y ); */
/* fy-= fac1; */
}
else /* first field */
fy += 0.5f/( (float)ibuf->y );
}
}
// pixel coordinates
minx = MIN3(dxt[0], dyt[0], dxt[0] + dyt[0]);
maxx = MAX3(dxt[0], dyt[0], dxt[0] + dyt[0]);
miny = MIN3(dxt[1], dyt[1], dxt[1] + dyt[1]);
maxy = MAX3(dxt[1], dyt[1], dxt[1] + dyt[1]);
// tex_sharper has been removed
minx = (maxx - minx)*0.5f;
miny = (maxy - miny)*0.5f;
if (tex->imaflag & TEX_FILTER_MIN) {
// make sure the filtersize is minimal in pixels (normal, ref map can have miniature pixel dx/dy)
spaces -> tabs, (4 spaces == 1 tab, only for white space preceding text)
2010-03-22 09:30:00 +00:00
const float addval = (0.5f * tex->filtersize) / (float)MIN2(ibuf->x, ibuf->y);
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
if (addval > minx) minx = addval;
if (addval > miny) miny = addval;
}
else if (tex->filtersize != 1.f) {
minx *= tex->filtersize;
miny *= tex->filtersize;
dxt[0] *= tex->filtersize;
dxt[1] *= tex->filtersize;
dyt[0] *= tex->filtersize;
dyt[1] *= tex->filtersize;
}
if (tex->imaflag & TEX_IMAROT) {
float t;
SWAP(float, minx, miny);
// must rotate dxt/dyt 90 deg
// yet another blender problem is that swapping X/Y axes (or any tex proj switches) should do something similar,
// but it doesn't, it only swaps coords, so filter area will be incorrect in those cases.
t = dxt[0];
dxt[0] = dxt[1];
dxt[1] = -t;
t = dyt[0];
dyt[0] = dyt[1];
dyt[1] = -t;
}
// side faces of unit-cube
minx = (minx > 0.25f) ? 0.25f : ((minx < 1e-5f) ? 1e-5 : minx);
miny = (miny > 0.25f) ? 0.25f : ((miny < 1e-5f) ? 1e-5 : miny);
// repeat and clip
if (tex->extend == TEX_REPEAT) {
if ((tex->flag & (TEX_REPEAT_XMIR | TEX_REPEAT_YMIR)) == (TEX_REPEAT_XMIR | TEX_REPEAT_YMIR))
extflag = TXC_EXTD;
else if (tex->flag & TEX_REPEAT_XMIR)
extflag = TXC_XMIR;
else if (tex->flag & TEX_REPEAT_YMIR)
extflag = TXC_YMIR;
else
extflag = TXC_REPT;
}
else if (tex->extend == TEX_EXTEND)
extflag = TXC_EXTD;
if (tex->extend == TEX_CHECKER) {
int xs = (int)floorf(fx), ys = (int)floorf(fy);
// both checkers available, no boundary exceptions, checkerdist will eat aliasing
if ((tex->flag & TEX_CHECKER_ODD) && (tex->flag & TEX_CHECKER_EVEN)) {
fx -= xs;
fy -= ys;
}
else {
int xs1 = (int)floorf(fx - minx);
int ys1 = (int)floorf(fy - miny);
int xs2 = (int)floorf(fx + minx);
int ys2 = (int)floorf(fy + miny);
if ((xs1 != xs2) || (ys1 != ys2)) {
if (tex->flag & TEX_CHECKER_ODD) {
fx -= ((xs1 + ys) & 1) ? xs2 : xs1;
fy -= ((ys1 + xs) & 1) ? ys2 : ys1;
}
if (tex->flag & TEX_CHECKER_EVEN) {
fx -= ((xs1 + ys) & 1) ? xs1 : xs2;
fy -= ((ys1 + xs) & 1) ? ys1 : ys2;
}
}
else {
if ((tex->flag & TEX_CHECKER_ODD) == 0 && ((xs + ys) & 1) == 0) return retval;
if ((tex->flag & TEX_CHECKER_EVEN) == 0 && (xs + ys) & 1) return retval;
fx -= xs;
fy -= ys;
}
}
// scale around center, (0.5, 0.5)
if (tex->checkerdist < 1.f) {
const float omcd = 1.f / (1.f - tex->checkerdist);
fx = (fx - 0.5f)*omcd + 0.5f;
fy = (fy - 0.5f)*omcd + 0.5f;
minx *= omcd;
miny *= omcd;
}
}
if (tex->extend == TEX_CLIPCUBE) {
if ((fx + minx) < 0.f || (fy + miny) < 0.f || (fx - minx) > 1.f || (fy - miny) > 1.f || texvec[2] < -1.f || texvec[2] > 1.f) return retval;
}
else if (tex->extend == TEX_CLIP || tex->extend == TEX_CHECKER) {
if ((fx + minx) < 0.f || (fy + miny) < 0.f || (fx - minx) > 1.f || (fy - miny) > 1.f) return retval;
}
else {
if (tex->extend == TEX_EXTEND) {
fx = (fx > 1.f) ? 1.f : ((fx < 0.f) ? 0.f : fx);
fy = (fy > 1.f) ? 1.f : ((fy < 0.f) ? 0.f : fy);
}
else {
fx -= floorf(fx);
fy -= floorf(fy);
}
}
intpol = tex->imaflag & TEX_INTERPOL;
// warning no return!
if ((R.flag & R_SEC_FIELD) && (ibuf->flags & IB_fields))
ibuf->rect += ibuf->x*ibuf->y;
// struct common data
Fix very long render time when using a large filter size with the new image sample code. I've added a dumb clamping now of dxt/dyt, probably could use a smarter solution but it's not likely to be noticeable.
2010-02-15 08:29:01 +00:00
copy_v2_v2(AFD.dxt, dxt);
copy_v2_v2(AFD.dyt, dyt);
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
AFD.intpol = intpol;
AFD.extflag = extflag;
Fix very long render time when using a large filter size with the new image sample code. I've added a dumb clamping now of dxt/dyt, probably could use a smarter solution but it's not likely to be noticeable.
2010-02-15 08:29:01 +00:00
// brecht: added stupid clamping here, large dx/dy can give very large
// filter sizes which take ages to render, it may be better to do this
// more intelligently later in the code .. probably it's not noticeable
if(AFD.dxt[0]*AFD.dxt[0] + AFD.dxt[1]*AFD.dxt[1] > 2.0f*2.0f) {
mul_v2_fl(AFD.dxt, 2.0f/len_v2(AFD.dxt));
mul_v2_fl(AFD.dyt, 2.0f/len_v2(AFD.dyt));
}
if(AFD.dyt[0]*AFD.dyt[0] + AFD.dyt[1]*AFD.dyt[1] > 2.0f*2.0f)
mul_v2_fl(AFD.dyt, 2.0f/len_v2(AFD.dyt));
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
// choice:
if (tex->imaflag & TEX_MIPMAP) {
ImBuf *previbuf, *curibuf;
float levf;
int maxlev;
ImBuf* mipmaps[IB_MIPMAP_LEVELS + 1];
// modify ellipse minor axis if too eccentric, use for area sampling as well
// scaling dxt/dyt as done in pbrt is not the same
// (as in ewa_eval(), scale by sqrt(ibuf->x) to maximize precision)
const float ff = sqrtf(ibuf->x), q = ibuf->y/ff;
const float Ux = dxt[0]*ff, Vx = dxt[1]*q, Uy = dyt[0]*ff, Vy = dyt[1]*q;
const float A = Vx*Vx + Vy*Vy;
const float B = -2.f*(Ux*Vx + Uy*Vy);
const float C = Ux*Ux + Uy*Uy;
const float F = A*C - B*B*0.25f;
float a, b, th, ecc;
imp2radangle(A, B, C, F, &a, &b, &th, &ecc);
if (tex->texfilter == TXF_FELINE) {
float fProbes;
a *= ff;
b *= ff;
a = MAX2(a, 1.f);
b = MAX2(b, 1.f);
fProbes = 2.f*(a / b) - 1.f;
AFD.iProbes = (int)floorf(fProbes + 0.5f);
AFD.iProbes = MIN2(AFD.iProbes, tex->afmax);
if (AFD.iProbes < fProbes)
b = 2.f*a / (float)(AFD.iProbes + 1);
AFD.majrad = a/ff;
AFD.minrad = b/ff;
AFD.theta = th;
AFD.dusc = 1.f/ff;
AFD.dvsc = ff / (float)ibuf->y;
}
else { // EWA & area
if (ecc > (float)tex->afmax) b = a / (float)tex->afmax;
b *= ff;
}
maxd = MAX2(b, 1e-8f);
2.5 * Make EWA new default, rename Default to Box. * Fix windows compile issue in texture filter code.
2009-07-22 17:20:04 +00:00
levf = ((float)M_LOG2E)*logf(maxd);
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
curmap = 0;
maxlev = 1;
mipmaps[0] = ibuf;
while (curmap < IB_MIPMAP_LEVELS) {
mipmaps[curmap + 1] = ibuf->mipmap[curmap];
if (ibuf->mipmap[curmap]) maxlev++;
curmap++;
}
// mipmap level
if (levf < 0.f) { // original image only
previbuf = curibuf = mipmaps[0];
levf = 0.f;
}
else if (levf >= maxlev - 1) {
previbuf = curibuf = mipmaps[maxlev - 1];
levf = 0.f;
if (tex->texfilter == TXF_FELINE) AFD.iProbes = 1;
}
else {
const int lev = ISNAN(levf) ? 0 : (int)levf;
curibuf = mipmaps[lev];
previbuf = mipmaps[lev + 1];
levf -= floorf(levf);
}
// filter functions take care of interpolation themselves, no need to modify dxt/dyt here
if (texres->nor && ((tex->imaflag & TEX_NORMALMAP) == 0)) {
// color & normal
filterfunc(texres, curibuf, fx, fy, &AFD);
val1 = texres->tr + texres->tg + texres->tb;
filterfunc(&texr, curibuf, fx + dxt[0], fy + dxt[1], &AFD);
val2 = texr.tr + texr.tg + texr.tb;
filterfunc(&texr, curibuf, fx + dyt[0], fy + dyt[1], &AFD);
val3 = texr.tr + texr.tg + texr.tb;
// don't switch x or y!
texres->nor[0] = val1 - val2;
texres->nor[1] = val1 - val3;
if (previbuf != curibuf) { // interpolate
filterfunc(&texr, previbuf, fx, fy, &AFD);
// rgb
texres->tr += levf*(texr.tr - texres->tr);
texres->tg += levf*(texr.tg - texres->tg);
texres->tb += levf*(texr.tb - texres->tb);
texres->ta += levf*(texr.ta - texres->ta);
// normal
val1 += levf*((texr.tr + texr.tg + texr.tb) - val1);
filterfunc(&texr, previbuf, fx + dxt[0], fy + dxt[1], &AFD);
val2 += levf*((texr.tr + texr.tg + texr.tb) - val2);
filterfunc(&texr, previbuf, fx + dyt[0], fy + dyt[1], &AFD);
val3 += levf*((texr.tr + texr.tg + texr.tb) - val3);
texres->nor[0] = val1 - val2; // vals have been interpolated above!
texres->nor[1] = val1 - val3;
}
}
else { // color
filterfunc(texres, curibuf, fx, fy, &AFD);
if (previbuf != curibuf) { // interpolate
filterfunc(&texr, previbuf, fx, fy, &AFD);
texres->tr += levf*(texr.tr - texres->tr);
texres->tg += levf*(texr.tg - texres->tg);
texres->tb += levf*(texr.tb - texres->tb);
texres->ta += levf*(texr.ta - texres->ta);
}
alpha_clip_aniso(ibuf, fx-minx, fy-miny, fx+minx, fy+miny, extflag, texres);
}
}
else { // no mipmap
// filter functions take care of interpolation themselves, no need to modify dxt/dyt here
if (tex->texfilter == TXF_FELINE) {
const float ff = sqrtf(ibuf->x), q = ibuf->y/ff;
const float Ux = dxt[0]*ff, Vx = dxt[1]*q, Uy = dyt[0]*ff, Vy = dyt[1]*q;
const float A = Vx*Vx + Vy*Vy;
const float B = -2.f*(Ux*Vx + Uy*Vy);
const float C = Ux*Ux + Uy*Uy;
const float F = A*C - B*B*0.25f;
float a, b, th, ecc, fProbes;
imp2radangle(A, B, C, F, &a, &b, &th, &ecc);
a *= ff;
b *= ff;
a = MAX2(a, 1.f);
b = MAX2(b, 1.f);
fProbes = 2.f*(a / b) - 1.f;
// no limit to number of Probes here
AFD.iProbes = (int)floorf(fProbes + 0.5f);
if (AFD.iProbes < fProbes) b = 2.f*a / (float)(AFD.iProbes + 1);
AFD.majrad = a/ff;
AFD.minrad = b/ff;
AFD.theta = th;
AFD.dusc = 1.f/ff;
AFD.dvsc = ff / (float)ibuf->y;
}
if (texres->nor && ((tex->imaflag & TEX_NORMALMAP) == 0)) {
// color & normal
filterfunc(texres, ibuf, fx, fy, &AFD);
val1 = texres->tr + texres->tg + texres->tb;
filterfunc(&texr, ibuf, fx + dxt[0], fy + dxt[1], &AFD);
val2 = texr.tr + texr.tg + texr.tb;
filterfunc(&texr, ibuf, fx + dyt[0], fy + dyt[1], &AFD);
val3 = texr.tr + texr.tg + texr.tb;
// don't switch x or y!
texres->nor[0] = val1 - val2;
texres->nor[1] = val1 - val3;
}
else {
filterfunc(texres, ibuf, fx, fy, &AFD);
alpha_clip_aniso(ibuf, fx-minx, fy-miny, fx+minx, fy+miny, extflag, texres);
}
}
if (tex->imaflag & TEX_CALCALPHA)
texres->ta = texres->tin = texres->ta * MAX3(texres->tr, texres->tg, texres->tb);
else
texres->tin = texres->ta;
if (tex->flag & TEX_NEGALPHA) texres->ta = 1.f - texres->ta;
if ((R.flag & R_SEC_FIELD) && (ibuf->flags & IB_fields))
ibuf->rect -= ibuf->x*ibuf->y;
if (texres->nor && (tex->imaflag & TEX_NORMALMAP)) { // normal from color
texres->nor[0] = 2.f*(texres->tr - 0.5f);
texres->nor[1] = 2.f*(0.5f - texres->tg);
texres->nor[2] = 2.f*(texres->tb - 0.5f);
}
// de-premul, this is being premulled in shade_input_do_shade()
// TXF: this currently does not (yet?) work properly, destroys edge AA in clip/checker mode, so for now commented out
// also disabled in imagewraposa() to be able to compare results with blender's default texture filtering
// brecht: tried to fix this, see "TXF alpha" comments
if (texres->ta != 1.f && (texres->ta > FLT_EPSILON)) {
fx = 1.f/texres->ta;
texres->tr *= fx;
texres->tg *= fx;
texres->tb *= fx;
}
Fix [#22610] Alpha problem with textureswhen Brightness > 1 or Contrast < 1 * Enabled premultiplication for packed images * Added pack/unpack operator to image template * Moved brightness/contrast corrections to after de-premultiplication in image texture sampling
2010-06-17 07:20:12 +00:00
BRICONTRGB;
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
return retval;
}
Composite & Pass render goodies: - New Passes: UV and Rad(iosity) - New Nodes: UV Map and Index Mask - Z-combine now is antialiased As usual, please check the log. Has nice pics! http://www.blender3d.org/cms/Composite__UV_Map__ID.830.0.html For devs: the antialias code from Vector Blur is now exported in compo too. Works pretty good. Even fixed a bug in antialias, so vectorblur will be better. Also: found out that OpenGL display list speedup accidentally was still triggered with the rt button... so it did not work by default.
2006-12-10 20:30:15 +00:00
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
int imagewraposa(Tex *tex, Image *ima, ImBuf *ibuf, float *texvec, float *DXT, float *DYT, TexResult *texres)
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
{
TexResult texr;
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
float fx, fy, minx, maxx, miny, maxy, dx, dy, dxt[3], dyt[3];
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 maxd, pixsize, val1, val2, val3;
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
int curmap, retval, imaprepeat, imapextend;
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
// TXF: since dxt/dyt might be modified here and since they might be needed after imagewraposa() call,
// make a local copy here so that original vecs remain untouched
VECCOPY(dxt, DXT);
VECCOPY(dyt, DYT);
// anisotropic filtering
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
if (tex->texfilter != TXF_BOX)
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
return imagewraposa_aniso(tex, ima, ibuf, texvec, dxt, dyt, texres);
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
texres->tin= texres->ta= texres->tr= texres->tg= texres->tb= 0.0f;
Bugfix #3608 More bump issues... Env made very nice examples, but all of it quite nasty to solve. Three improvements I found nevertheless: - bug in mipmap code for bump gave errors in one derivate, causing aliasing - when Image texture option "Clip" was used, the imagetexture function returned wrong value, causing the caller to create own normal based on color (zero!). - Mipmapped bump now also interpolates the samples (never did this...)
2005-12-19 14:00:32 +00:00
/* we need to set retval OK, otherwise texture code generates normals itself... */
retval= texres->nor?3:1;
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* quick tests */
if(ibuf==NULL && ima==NULL)
Bugfix #3608 More bump issues... Env made very nice examples, but all of it quite nasty to solve. Three improvements I found nevertheless: - bug in mipmap code for bump gave errors in one derivate, causing aliasing - when Image texture option "Clip" was used, the imagetexture function returned wrong value, causing the caller to create own normal based on color (zero!). - Mipmapped bump now also interpolates the samples (never did this...)
2005-12-19 14:00:32 +00:00
return retval;
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(ima) {
/* hack for icon render */
if(ima->ibufs.first==NULL && (R.r.scemode & R_NO_IMAGE_LOAD))
return retval;
ibuf= BKE_image_get_ibuf(ima, &tex->iuser);
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(ibuf==NULL || (ibuf->rect==NULL && ibuf->rect_float==NULL))
return retval;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* mipmap test */
Bugfix #22040 Old bug report: Image Editor, Painting: crash when texture was visible in Material or Texture preview. Was caused by paint code freeing mipmaps. Now replaced with a mipmap tag (to be done again), and a new mipmap function that doesn't re-allocate.
2010-12-14 18:02:41 +00:00
image_mipmap_test(tex, ibuf);
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->imaflag & TEX_USEALPHA) {
if(tex->imaflag & TEX_CALCALPHA);
else texres->talpha= 1;
Fix for bug #5044: random crashes when rendering. This one is of my own making, removed too many locks. Thanks Malefico for reporting.
2006-09-30 21:29:47 +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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
texr.talpha= texres->talpha;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->imaflag & TEX_IMAROT) {
fy= texvec[0];
fx= texvec[1];
}
else {
fx= texvec[0];
fy= texvec[1];
}
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(ibuf->flags & IB_fields) {
if(R.r.mode & R_FIELDS) { /* field render */
if(R.flag & R_SEC_FIELD) { /* correction for 2nd field */
/* fac1= 0.5/( (float)ibuf->y ); */
/* fy-= fac1; */
}
else { /* first field */
fy+= 0.5f/( (float)ibuf->y );
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
}
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
}
/* pixel coordinates */
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
minx= MIN3(dxt[0],dyt[0],dxt[0]+dyt[0] );
maxx= MAX3(dxt[0],dyt[0],dxt[0]+dyt[0] );
miny= MIN3(dxt[1],dyt[1],dxt[1]+dyt[1] );
maxy= MAX3(dxt[1],dyt[1],dxt[1]+dyt[1] );
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* tex_sharper has been removed */
Error in previous commit: the texture filter size should not be scaled at all if you use new 'minimal' option. The new option works much nicer if it only sets the minimum. (use texture with 'refl' map) Added option to env map too!
2007-12-28 19:38:47 +00:00
minx= (maxx-minx)/2.0f;
miny= (maxy-miny)/2.0f;
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
And there's another annoyance I got poked for: Image texture "Filter size" was not well usable for making the appearance soft filtered, this because it multiplied the sample values, and such values could be extreme small. Added next to "Filter" buton a new "Min" option, which enforces a filter size to be a minimum of 'filter' pixels in size.
2007-12-28 18:58:40 +00:00
if(tex->imaflag & TEX_FILTER_MIN) {
/* make sure the filtersize is minimal in pixels (normal, ref map can have miniature pixel dx/dy) */
spaces -> tabs, (4 spaces == 1 tab, only for white space preceding text)
2010-03-22 09:30:00 +00:00
float addval= (0.5f * tex->filtersize) / (float) MIN2(ibuf->x, ibuf->y);
And there's another annoyance I got poked for: Image texture "Filter size" was not well usable for making the appearance soft filtered, this because it multiplied the sample values, and such values could be extreme small. Added next to "Filter" buton a new "Min" option, which enforces a filter size to be a minimum of 'filter' pixels in size.
2007-12-28 18:58:40 +00:00
if(addval > minx)
minx= addval;
if(addval > miny)
miny= addval;
}
Error in previous commit: the texture filter size should not be scaled at all if you use new 'minimal' option. The new option works much nicer if it only sets the minimum. (use texture with 'refl' map) Added option to env map too!
2007-12-28 19:38:47 +00:00
else if(tex->filtersize!=1.0f) {
minx*= tex->filtersize;
miny*= tex->filtersize;
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
dxt[0]*= tex->filtersize;
dxt[1]*= tex->filtersize;
dyt[0]*= tex->filtersize;
dyt[1]*= tex->filtersize;
}
Bugfix #2841 Ancient issue with Image bumpmap & orientation... this report was about UV textures, which flipped bump direction when UV coordinates rotate. So... time to dive deeper into it, and I found a very bad old construction in the way image textures handled bump normals. Instead of using the correct dx,dy vectors it was calculating the boundbox for it, and applied it as such (loosing negative direction for the dx or dy that way). Anyhoo! To make a long story short; the weird "correct Nor Map" option now can be removed... but requires careful testing! Here's renders from file I got from env, 239 means old and 240 is this cvs. http://www.blender.org/bf/dino239.jpg http://www.blender.org/bf/dino240.jpg http://www.blender.org/bf/dino239a.jpg http://www.blender.org/bf/dino240a.jpg Note the much finer detail, and the more consistant normals direction. No, go render testing dudes!
2005-11-21 22:47:23 +00:00
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->imaflag & TEX_IMAROT) SWAP(float, minx, miny);
if(minx>0.25) minx= 0.25;
else if(minx<0.00001f) minx= 0.00001f; /* side faces of unit-cube */
if(miny>0.25) miny= 0.25;
else if(miny<0.00001f) miny= 0.00001f;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* repeat and clip */
imaprepeat= (tex->extend==TEX_REPEAT);
imapextend= (tex->extend==TEX_EXTEND);
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
Important bugfix: Image mapping "repeat" didn't support mirrored tiling yet: http://www.blender.org/bf/rep1.jpg Psst psst... to solve this bug I had to add the buttons for the option too!
2007-01-14 11:51:52 +00:00
if(tex->extend == TEX_REPEAT) {
if(tex->flag & (TEX_REPEAT_XMIR|TEX_REPEAT_YMIR)) {
imaprepeat= 0;
imapextend= 1;
}
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->extend == TEX_CHECKER) {
int xs, ys, xs1, ys1, xs2, ys2, boundary;
xs= (int)floor(fx);
ys= (int)floor(fy);
// both checkers available, no boundary exceptions, checkerdist will eat aliasing
if( (tex->flag & TEX_CHECKER_ODD) && (tex->flag & TEX_CHECKER_EVEN) ) {
fx-= xs;
fy-= ys;
}
else {
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
xs1= (int)floor(fx-minx);
ys1= (int)floor(fy-miny);
xs2= (int)floor(fx+minx);
ys2= (int)floor(fy+miny);
boundary= (xs1!=xs2) || (ys1!=ys2);
if(boundary==0) {
if( (tex->flag & TEX_CHECKER_ODD)==0) {
if((xs+ys) & 1);
else return retval;
}
if( (tex->flag & TEX_CHECKER_EVEN)==0) {
if((xs+ys) & 1) return retval;
}
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
fx-= xs;
fy-= ys;
}
else {
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->flag & TEX_CHECKER_ODD) {
if((xs1+ys) & 1) fx-= xs2;
else fx-= xs1;
if((ys1+xs) & 1) fy-= ys2;
else fy-= ys1;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->flag & TEX_CHECKER_EVEN) {
if((xs1+ys) & 1) fx-= xs1;
else fx-= xs2;
if((ys1+xs) & 1) fy-= ys1;
else fy-= ys2;
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
}
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* scale around center, (0.5, 0.5) */
if(tex->checkerdist<1.0) {
fx= (fx-0.5)/(1.0-tex->checkerdist) +0.5;
fy= (fy-0.5)/(1.0-tex->checkerdist) +0.5;
minx/= (1.0-tex->checkerdist);
miny/= (1.0-tex->checkerdist);
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->extend == TEX_CLIPCUBE) {
if(fx+minx<0.0 || fy+miny<0.0 || fx-minx>1.0 || fy-miny>1.0 || texvec[2]<-1.0 || texvec[2]>1.0) {
return retval;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
}
else if(tex->extend==TEX_CLIP || tex->extend==TEX_CHECKER) {
if(fx+minx<0.0 || fy+miny<0.0 || fx-minx>1.0 || fy-miny>1.0) {
return retval;
}
}
else {
Important bugfix: Image mapping "repeat" didn't support mirrored tiling yet: http://www.blender.org/bf/rep1.jpg Psst psst... to solve this bug I had to add the buttons for the option too!
2007-01-14 11:51:52 +00:00
if(imapextend) {
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(fx>1.0) fx = 1.0;
else if(fx<0.0) fx= 0.0;
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
}
else {
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(fx>1.0) fx -= (int)(fx);
else if(fx<0.0) fx+= 1-(int)(fx);
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
Important bugfix: Image mapping "repeat" didn't support mirrored tiling yet: http://www.blender.org/bf/rep1.jpg Psst psst... to solve this bug I had to add the buttons for the option too!
2007-01-14 11:51:52 +00:00
if(imapextend) {
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(fy>1.0) fy = 1.0;
else if(fy<0.0) fy= 0.0;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
else {
if(fy>1.0) fy -= (int)(fy);
else if(fy<0.0) fy+= 1-(int)(fy);
}
}
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* warning no return! */
if( (R.flag & R_SEC_FIELD) && (ibuf->flags & IB_fields) ) {
ibuf->rect+= (ibuf->x*ibuf->y);
}
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* choice: */
if(tex->imaflag & TEX_MIPMAP) {
ImBuf *previbuf, *curibuf;
float bumpscale;
dx= minx;
dy= miny;
maxd= MAX2(dx, dy);
if(maxd>0.5) maxd= 0.5;
Interesting commit for artists using huge textures; The code that generated mipmaps took a real long time to do it... on a 5k x 5k image it took here (no optim, debug compile) 32.5 sec. Recoded the very old filtering routine, which already brought it down to 2.8 seconds. Then tested if we even need this filtering... in many cases the images are painted or photographs, which is filtered OK already. Without the filter, the mipmap timing went down to 0.39 second. :) http://www.blender.org/bf/filters/index1.html Here's an example of two 'mips' generated with or without gauss filter. Note that aliasing in an image remains there... which can be a wanted effect anyway. So; added the gauss filter as option in making mipmaps. Also had to reshuffle the buttons there in a more logical manner. There's also disabled code in the do_versions to set 'gauss' on in older files. Will be enabled during release time.
2006-02-10 18:57:52 +00:00
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
pixsize = 1.0f/ (float) MIN2(ibuf->x, ibuf->y);
bumpscale= pixsize/maxd;
if(bumpscale>1.0f) bumpscale= 1.0f;
else bumpscale*=bumpscale;
curmap= 0;
previbuf= curibuf= ibuf;
while(curmap<IB_MIPMAP_LEVELS && ibuf->mipmap[curmap]) {
if(maxd < pixsize) break;
previbuf= curibuf;
curibuf= ibuf->mipmap[curmap];
pixsize= 1.0f / (float)MIN2(curibuf->x, curibuf->y);
curmap++;
}
if(previbuf!=curibuf || (tex->imaflag & TEX_INTERPOL)) {
/* sample at least 1 pixel */
if (minx < 0.5f / ibuf->x) minx = 0.5f / ibuf->x;
if (miny < 0.5f / ibuf->y) miny = 0.5f / ibuf->y;
}
if(texres->nor && (tex->imaflag & TEX_NORMALMAP)==0) {
/* a bit extra filter */
//minx*= 1.35f;
//miny*= 1.35f;
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
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(curibuf, fx-minx, fy-miny, fx+minx, fy+miny, texres, imaprepeat, imapextend);
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
val1= texres->tr+texres->tg+texres->tb;
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(curibuf, fx-minx+dxt[0], fy-miny+dxt[1], fx+minx+dxt[0], fy+miny+dxt[1], &texr, imaprepeat, imapextend);
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
val2= texr.tr + texr.tg + texr.tb;
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(curibuf, fx-minx+dyt[0], fy-miny+dyt[1], fx+minx+dyt[0], fy+miny+dyt[1], &texr, imaprepeat, imapextend);
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
val3= texr.tr + texr.tg + texr.tb;
/* don't switch x or y! */
texres->nor[0]= (val1-val2);
texres->nor[1]= (val1-val3);
if(previbuf!=curibuf) { /* interpolate */
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
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(previbuf, fx-minx, fy-miny, fx+minx, fy+miny, &texr, imaprepeat, imapextend);
Bugfix #3608 More bump issues... Env made very nice examples, but all of it quite nasty to solve. Three improvements I found nevertheless: - bug in mipmap code for bump gave errors in one derivate, causing aliasing - when Image texture option "Clip" was used, the imagetexture function returned wrong value, causing the caller to create own normal based on color (zero!). - Mipmapped bump now also interpolates the samples (never did this...)
2005-12-19 14:00:32 +00:00
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* calc rgb */
dx= 2.0f*(pixsize-maxd)/pixsize;
if(dx>=1.0f) {
texres->ta= texr.ta; texres->tb= texr.tb;
texres->tg= texr.tg; texres->tr= texr.tr;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
else {
dy= 1.0f-dx;
texres->tb= dy*texres->tb+ dx*texr.tb;
texres->tg= dy*texres->tg+ dx*texr.tg;
texres->tr= dy*texres->tr+ dx*texr.tr;
texres->ta= dy*texres->ta+ dx*texr.ta;
}
val1= dy*val1+ dx*(texr.tr + texr.tg + texr.tb);
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(previbuf, fx-minx+dxt[0], fy-miny+dxt[1], fx+minx+dxt[0], fy+miny+dxt[1], &texr, imaprepeat, imapextend);
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
val2= dy*val2+ dx*(texr.tr + texr.tg + texr.tb);
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(previbuf, fx-minx+dyt[0], fy-miny+dyt[1], fx+minx+dyt[0], fy+miny+dyt[1], &texr, imaprepeat, imapextend);
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
val3= dy*val3+ dx*(texr.tr + texr.tg + texr.tb);
texres->nor[0]= (val1-val2); /* vals have been interpolated above! */
texres->nor[1]= (val1-val3);
if(dx<1.0f) {
dy= 1.0f-dx;
texres->tb= dy*texres->tb+ dx*texr.tb;
texres->tg= dy*texres->tg+ dx*texr.tg;
texres->tr= dy*texres->tr+ dx*texr.tr;
texres->ta= dy*texres->ta+ dx*texr.ta;
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
}
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
texres->nor[0]*= bumpscale;
texres->nor[1]*= bumpscale;
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
}
else {
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
maxx= fx+minx;
minx= fx-minx;
maxy= fy+miny;
miny= fy-miny;
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(curibuf, minx, miny, maxx, maxy, texres, imaprepeat, imapextend);
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(previbuf!=curibuf) { /* interpolate */
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(previbuf, minx, miny, maxx, maxy, &texr, imaprepeat, imapextend);
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
fx= 2.0f*(pixsize-maxd)/pixsize;
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
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(fx>=1.0) {
texres->ta= texr.ta; texres->tb= texr.tb;
texres->tg= texr.tg; texres->tr= texr.tr;
} else {
fy= 1.0f-fx;
texres->tb= fy*texres->tb+ fx*texr.tb;
texres->tg= fy*texres->tg+ fx*texr.tg;
texres->tr= fy*texres->tr+ fx*texr.tr;
texres->ta= fy*texres->ta+ fx*texr.ta;
}
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
}
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
}
else {
2.5: Texture Filtering Patch by Alfredo de Greef with high quality image texture filters. This adds 3 new filters: * SAT: Summed Area Tables. This is like mipmaps, but using somewhat more memory avoids some artifacts. * EWA: Ellipitical Weighted Average, anisotropic filter. * FELINE: Fast elliptical lines for anisotropic texture mapping. The one change I made to this was to try to fix an alpha/premul problem, hopefully I didn't break anything, it looks compatible with the existing filter now for me.
2009-07-21 13:20:35 +00:00
const int intpol = tex->imaflag & TEX_INTERPOL;
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
if (intpol) {
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* sample 1 pixel minimum */
if (minx < 0.5f / ibuf->x) minx = 0.5f / ibuf->x;
if (miny < 0.5f / ibuf->y) miny = 0.5f / ibuf->y;
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(texres->nor && (tex->imaflag & TEX_NORMALMAP)==0) {
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(ibuf, fx-minx, fy-miny, fx+minx, fy+miny, texres, imaprepeat, imapextend);
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
val1= texres->tr+texres->tg+texres->tb;
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(ibuf, fx-minx+dxt[0], fy-miny+dxt[1], fx+minx+dxt[0], fy+miny+dxt[1], &texr, imaprepeat, imapextend);
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
val2= texr.tr + texr.tg + texr.tb;
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(ibuf, fx-minx+dyt[0], fy-miny+dyt[1], fx+minx+dyt[0], fy+miny+dyt[1], &texr, imaprepeat, imapextend);
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
val3= texr.tr + texr.tg + texr.tb;
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
/* don't switch x or y! */
texres->nor[0]= (val1-val2);
texres->nor[1]= (val1-val3);
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
}
Remove SAT texture filter. It's not working, thought it was but that's because the mipmap was not being refreshed. Also this will be problematic to support when I add tile/mipmap cache, so would not rather not try to. Can be added back afterwards if someone wants to make it work.
2010-03-14 18:22:04 +00:00
else
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(ibuf, fx-minx, fy-miny, fx+minx, fy+miny, texres, imaprepeat, imapextend);
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
}
The Big Image refactor! Please read: http://www.blender3d.org/cms/Imaging.834.0.html Or in short: - adding MultiLayer Image support - recoded entire Image API - better integration of movie/sequence Images Was a whole load of work... went down for a week to do this. So, will need a lot of testing! Will be in irc all evening.
2006-12-20 17:57:56 +00:00
if(tex->imaflag & TEX_CALCALPHA) {
texres->ta= texres->tin= texres->ta*MAX3(texres->tr, texres->tg, texres->tb);
}
else texres->tin= texres->ta;
if(tex->flag & TEX_NEGALPHA) texres->ta= 1.0f-texres->ta;
if( (R.flag & R_SEC_FIELD) && (ibuf->flags & IB_fields) ) {
ibuf->rect-= (ibuf->x*ibuf->y);
}
if(texres->nor && (tex->imaflag & TEX_NORMALMAP)) {
// qdn: normal from color
texres->nor[0] = 2.f*(texres->tr - 0.5f);
texres->nor[1] = 2.f*(0.5f - texres->tg);
texres->nor[2] = 2.f*(texres->tb - 0.5f);
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
}
Bugfix #7802 Fixed very old annoyance in alpha render: when using alpha textures images, nicely premulled as Blender wants it, you still got dark outlines at the edges. This because the blender shading pipeline also premultiplies. Solution: de-premul image texture colors after sampling.
2007-11-23 14:48:26 +00:00
/* de-premul, this is being premulled in shade_input_do_shade() */
Fix for bug #8750: depremul in the image texture code should check with an epsilon rather than 0.0f, could give random spots otherwise with e.g. calcalpha + negalpha.
2008-03-26 15:25:56 +00:00
if(texres->ta!=1.0f && texres->ta>FLT_EPSILON) {
Bugfix #7802 Fixed very old annoyance in alpha render: when using alpha textures images, nicely premulled as Blender wants it, you still got dark outlines at the edges. This because the blender shading pipeline also premultiplies. Solution: de-premul image texture colors after sampling.
2007-11-23 14:48:26 +00:00
fx= 1.0f/texres->ta;
texres->tr*= fx;
texres->tg*= fx;
texres->tb*= fx;
}
Fix [#22610] Alpha problem with textureswhen Brightness > 1 or Contrast < 1 * Enabled premultiplication for packed images * Added pack/unpack operator to image template * Moved brightness/contrast corrections to after de-premultiplication in image texture sampling
2010-06-17 07:20:12 +00:00
BRICONTRGB;
Bugfix #3608 More bump issues... Env made very nice examples, but all of it quite nasty to solve. Three improvements I found nevertheless: - bug in mipmap code for bump gave errors in one derivate, causing aliasing - when Image texture option "Clip" was used, the imagetexture function returned wrong value, causing the caller to create own normal based on color (zero!). - Mipmapped bump now also interpolates the samples (never did this...)
2005-12-19 14:00:32 +00:00
return retval;
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
}
Fix [#22418] Displace Node also makes entire image fuzzy Changed displace node sampling to use EWA filtering, and removed old hacks for calculating derivatives - I think it should be generated correctly now.
2010-05-26 08:06:51 +00:00
void image_sample(Image *ima, float fx, float fy, float dx, float dy, float *result)
{
TexResult texres;
ImBuf *ibuf= BKE_image_get_ibuf(ima, NULL);
if(ibuf==NULL) {
result[0]= result[1]= result[2]= result[3]= 0.0f;
return;
}
if( (R.flag & R_SEC_FIELD) && (ibuf->flags & IB_fields) )
ibuf->rect+= (ibuf->x*ibuf->y);
remove/tag unused args for view*.c, gpu*.c & image*.c
2010-10-15 12:29:02 +00:00
boxsample(ibuf, fx, fy, fx+dx, fy+dy, &texres, 0, 1);
Fix [#22418] Displace Node also makes entire image fuzzy Changed displace node sampling to use EWA filtering, and removed old hacks for calculating derivatives - I think it should be generated correctly now.
2010-05-26 08:06:51 +00:00
result[0]= texres.tr;
result[1]= texres.tg;
result[2]= texres.tb;
result[3]= texres.ta;
if( (R.flag & R_SEC_FIELD) && (ibuf->flags & IB_fields) )
ibuf->rect-= (ibuf->x*ibuf->y);
}
void ibuf_sample(ImBuf *ibuf, float fx, float fy, float dx, float dy, float *result)
{
TexResult texres;
afdata_t AFD;
if(ibuf==NULL) {
return;
}
AFD.dxt[0] = dx; AFD.dxt[1] = dx;
AFD.dyt[0] = dy; AFD.dyt[1] = dy;
//copy_v2_v2(AFD.dxt, dx);
//copy_v2_v2(AFD.dyt, dy);
AFD.intpol = 1;
AFD.extflag = TXC_EXTD;
memset(&texres, 0, sizeof(texres));
ewa_eval(&texres, ibuf, fx, fy, &AFD);
result[0]= texres.tr;
result[1]= texres.tg;
result[2]= texres.tb;
result[3]= texres.ta;
SVN maintenance.
2010-09-03 03:30:20 +00:00
}
Copy Permalink