This repository has been archived on 2023-10-09.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
blender-archive/source/blender/blenkernel/intern/image.c

3668 lines
92 KiB
C
Raw Normal View History

2011-10-10 09:38:02 +00:00
/*
* ***** BEGIN GPL LICENSE BLOCK *****
2002-10-12 11:37:38 +00:00
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
2002-10-12 11:37:38 +00:00
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software Foundation,
2010-02-12 13:34:04 +00:00
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
2002-10-12 11:37:38 +00:00
*
* The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
* All rights reserved.
*
* Contributor(s): Blender Foundation, 2006, full recode
2002-10-12 11:37:38 +00:00
*
* ***** END GPL LICENSE BLOCK *****
2002-10-12 11:37:38 +00:00
*/
2011-02-27 20:40:57 +00:00
/** \file blender/blenkernel/intern/image.c
* \ingroup bke
*/
2002-10-12 11:37:38 +00:00
#include <stdio.h>
#include <string.h>
#include <fcntl.h>
#include <math.h>
2012-06-23 23:22:19 +00:00
#ifndef WIN32
# include <unistd.h>
2002-10-12 11:37:38 +00:00
#else
# include <io.h>
2002-10-12 11:37:38 +00:00
#endif
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
#include <time.h>
2002-10-12 11:37:38 +00:00
#include "MEM_guardedalloc.h"
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
#include "IMB_colormanagement.h"
2002-10-12 11:37:38 +00:00
#include "IMB_imbuf_types.h"
#include "IMB_imbuf.h"
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
#include "IMB_moviecache.h"
#ifdef WITH_OPENEXR
# include "intern/openexr/openexr_multi.h"
#endif
2002-10-12 11:37:38 +00:00
#include "DNA_packedFile_types.h"
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
#include "DNA_scene_types.h"
#include "DNA_object_types.h"
#include "DNA_camera_types.h"
#include "DNA_sequence_types.h"
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
#include "DNA_userdef_types.h"
#include "DNA_brush_types.h"
#include "DNA_mesh_types.h"
#include "DNA_meshdata_types.h"
2002-10-12 11:37:38 +00:00
#include "BLI_blenlib.h"
#include "BLI_math_vector.h"
#include "BLI_threads.h"
#include "BLI_timecode.h" /* for stamp timecode format */
#include "BLI_utildefines.h"
2002-10-12 11:37:38 +00:00
Biiig commit! Thanks to 2-3 weeks of cvs freeze... Render: - New; support for dual CPU render (SDL thread) Currently only works with alternating scanlines, but gives excellent performance. For both normal render as unified implemented. Note the "mutex" locks on z-transp buffer render and imbuf loads. - This has been made possible by major cleanups in render code, especially getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct OSA or using Materials or Texture data to write to. - Made normal render fully 4x32 floats too, and removed all old optimizes with chars or shorts. - Made normal render and unified render use same code for sky and halo render, giving equal (and better) results for halo render. Old render now also uses PostProcess options (brightness, mul, gamma) - Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer after render. Using PostProcess menu you will note an immediate re- display of image too (32 bits RGBA) - Added "Hue" and "Saturation" sliders to PostProcess options - Render module is still not having a "nice" API, but amount of dependencies went down a lot. Next todo: remove abusive "previewrender" code. The last main global in Render (struct Render) now can be re-used for fully controlling a render, to allow multiple "instances" of render to open. - Renderwindow now displays a smal bar on top with the stats, and keeps the stats after render too. Including "spare" page support. Not only easier visible that way, but also to remove the awkward code that was drawing stats in the Info header (extreme slow on some ATIs too) - Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping defines. - I might have forgotten stuff... and will write a nice doc on the architecture!
2004-12-27 19:28:52 +00:00
#include "BKE_bmfont.h"
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
#include "BKE_colortools.h"
2002-10-12 11:37:38 +00:00
#include "BKE_global.h"
#include "BKE_icons.h"
2002-10-12 11:37:38 +00:00
#include "BKE_image.h"
#include "BKE_library.h"
#include "BKE_main.h"
#include "BKE_packedFile.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_scene.h"
#include "BKE_node.h"
#include "BKE_sequencer.h" /* seq_foreground_frame_get() */
2002-10-12 11:37:38 +00:00
#include "BLF_api.h"
#include "PIL_time.h"
#include "RE_pipeline.h"
#include "GPU_draw.h"
#include "BLI_sys_types.h" // for intptr_t support
2008-08-17 17:12:10 +00:00
/* for image user iteration */
#include "DNA_node_types.h"
#include "DNA_space_types.h"
#include "DNA_screen_types.h"
#include "DNA_view3d_types.h"
#include "WM_api.h"
static SpinLock image_spin;
/* max int, to indicate we don't store sequences in ibuf */
#define IMA_NO_INDEX 0x7FEFEFEF
/* quick lookup: supports 1 million frames, thousand passes */
#define IMA_MAKE_INDEX(frame, index) ((frame) << 10) + index
#define IMA_INDEX_FRAME(index) (index >> 10)
2013-12-18 00:20:18 +11:00
/*
#define IMA_INDEX_PASS(index) (index & ~1023)
2013-12-18 00:20:18 +11:00
*/
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
/* ******** IMAGE CACHE ************* */
typedef struct ImageCacheKey {
int index;
} ImageCacheKey;
static unsigned int imagecache_hashhash(const void *key_v)
{
const ImageCacheKey *key = key_v;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
return key->index;
}
static bool imagecache_hashcmp(const void *a_v, const void *b_v)
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
{
const ImageCacheKey *a = a_v;
const ImageCacheKey *b = b_v;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
return (a->index != b->index);
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
}
static void imagecache_keydata(void *userkey, int *framenr, int *proxy, int *render_flags)
{
ImageCacheKey *key = userkey;
*framenr = IMA_INDEX_FRAME(key->index);
*proxy = IMB_PROXY_NONE;
*render_flags = 0;
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
static void imagecache_put(Image *image, int index, ImBuf *ibuf)
{
ImageCacheKey key;
if (image->cache == NULL) {
// char cache_name[64];
// BLI_snprintf(cache_name, sizeof(cache_name), "Image Datablock %s", image->id.name);
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
image->cache = IMB_moviecache_create("Image Datablock Cache", sizeof(ImageCacheKey),
imagecache_hashhash, imagecache_hashcmp);
IMB_moviecache_set_getdata_callback(image->cache, imagecache_keydata);
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
}
key.index = index;
IMB_moviecache_put(image->cache, &key, ibuf);
}
static struct ImBuf *imagecache_get(Image *image, int index)
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
{
if (image->cache) {
ImageCacheKey key;
key.index = index;
return IMB_moviecache_get(image->cache, &key);
}
return NULL;
}
void BKE_images_init(void)
{
BLI_spin_init(&image_spin);
}
void BKE_images_exit(void)
{
BLI_spin_end(&image_spin);
}
/* ******** IMAGE PROCESSING ************* */
static void de_interlace_ng(struct ImBuf *ibuf) /* neogeo fields */
{
struct ImBuf *tbuf1, *tbuf2;
2012-06-23 23:22:19 +00:00
if (ibuf == NULL) return;
if (ibuf->flags & IB_fields) return;
ibuf->flags |= IB_fields;
2012-06-23 23:22:19 +00:00
if (ibuf->rect) {
/* make copies */
tbuf1 = IMB_allocImBuf(ibuf->x, (ibuf->y >> 1), (unsigned char)32, (int)IB_rect);
tbuf2 = IMB_allocImBuf(ibuf->x, (ibuf->y >> 1), (unsigned char)32, (int)IB_rect);
2012-06-23 23:22:19 +00:00
ibuf->x *= 2;
2012-06-23 23:22:19 +00:00
IMB_rectcpy(tbuf1, ibuf, 0, 0, 0, 0, ibuf->x, ibuf->y);
IMB_rectcpy(tbuf2, ibuf, 0, 0, tbuf2->x, 0, ibuf->x, ibuf->y);
2012-06-23 23:22:19 +00:00
ibuf->x /= 2;
IMB_rectcpy(ibuf, tbuf1, 0, 0, 0, 0, tbuf1->x, tbuf1->y);
IMB_rectcpy(ibuf, tbuf2, 0, tbuf2->y, 0, 0, tbuf2->x, tbuf2->y);
2012-06-23 23:22:19 +00:00
IMB_freeImBuf(tbuf1);
IMB_freeImBuf(tbuf2);
2002-10-12 11:37:38 +00:00
}
ibuf->y /= 2;
}
static void de_interlace_st(struct ImBuf *ibuf) /* standard fields */
{
struct ImBuf *tbuf1, *tbuf2;
2012-06-23 23:22:19 +00:00
if (ibuf == NULL) return;
if (ibuf->flags & IB_fields) return;
ibuf->flags |= IB_fields;
2012-06-23 23:22:19 +00:00
if (ibuf->rect) {
/* make copies */
tbuf1 = IMB_allocImBuf(ibuf->x, (ibuf->y >> 1), (unsigned char)32, IB_rect);
tbuf2 = IMB_allocImBuf(ibuf->x, (ibuf->y >> 1), (unsigned char)32, IB_rect);
2012-06-23 23:22:19 +00:00
ibuf->x *= 2;
2012-06-23 23:22:19 +00:00
IMB_rectcpy(tbuf1, ibuf, 0, 0, 0, 0, ibuf->x, ibuf->y);
IMB_rectcpy(tbuf2, ibuf, 0, 0, tbuf2->x, 0, ibuf->x, ibuf->y);
2012-06-23 23:22:19 +00:00
ibuf->x /= 2;
IMB_rectcpy(ibuf, tbuf2, 0, 0, 0, 0, tbuf2->x, tbuf2->y);
IMB_rectcpy(ibuf, tbuf1, 0, tbuf2->y, 0, 0, tbuf1->x, tbuf1->y);
2012-06-23 23:22:19 +00:00
IMB_freeImBuf(tbuf1);
IMB_freeImBuf(tbuf2);
}
ibuf->y /= 2;
}
void BKE_image_de_interlace(Image *ima, int odd)
{
ImBuf *ibuf = BKE_image_acquire_ibuf(ima, NULL, NULL);
if (ibuf) {
if (odd)
de_interlace_st(ibuf);
else
de_interlace_ng(ibuf);
}
BKE_image_release_ibuf(ima, ibuf, NULL);
2002-10-12 11:37:38 +00:00
}
/* ***************** ALLOC & FREE, DATA MANAGING *************** */
2014-10-17 12:37:40 +02:00
static void image_free_cached_frames(Image *image)
{
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
if (image->cache) {
IMB_moviecache_free(image->cache);
image->cache = NULL;
}
}
/**
* Simply free the image data from memory,
* on display the image can load again (except for render buffers).
*/
2014-05-22 17:40:35 +10:00
void BKE_image_free_buffers(Image *ima)
{
2014-10-17 12:37:40 +02:00
image_free_cached_frames(ima);
2012-06-23 23:22:19 +00:00
if (ima->anim) IMB_free_anim(ima->anim);
ima->anim = NULL;
if (ima->rr) {
RE_FreeRenderResult(ima->rr);
ima->rr = NULL;
2012-06-23 23:22:19 +00:00
}
if (!G.background) {
/* Background mode doesn't use opnegl,
* so we can avoid freeing GPU images and save some
* time by skipping mutex lock.
*/
GPU_free_image(ima);
}
2012-06-23 23:22:19 +00:00
ima->ok = IMA_OK;
}
2002-10-12 11:37:38 +00:00
/* called by library too, do not free ima itself */
void BKE_image_free(Image *ima)
2002-10-12 11:37:38 +00:00
{
int a;
2014-05-22 17:40:35 +10:00
BKE_image_free_buffers(ima);
2002-10-12 11:37:38 +00:00
if (ima->packedfile) {
freePackedFile(ima->packedfile);
ima->packedfile = NULL;
}
BKE_icon_delete(&ima->id);
ima->id.icon_id = 0;
BKE_previewimg_free(&ima->preview);
for (a = 0; a < IMA_MAX_RENDER_SLOT; a++) {
if (ima->renders[a]) {
RE_FreeRenderResult(ima->renders[a]);
ima->renders[a] = NULL;
}
}
2002-10-12 11:37:38 +00:00
}
/* only image block itself */
static Image *image_alloc(Main *bmain, const char *name, short source, short type)
{
Image *ima;
2012-06-23 23:22:19 +00:00
ima = BKE_libblock_alloc(bmain, ID_IM, name);
if (ima) {
ima->ok = IMA_OK;
2012-06-23 23:22:19 +00:00
ima->xrep = ima->yrep = 1;
ima->aspx = ima->aspy = 1.0;
ima->gen_x = 1024; ima->gen_y = 1024;
ima->gen_type = 1; /* no defines yet? */
2012-06-23 23:22:19 +00:00
ima->source = source;
ima->type = type;
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
if (source == IMA_SRC_VIEWER)
ima->flag |= IMA_VIEW_AS_RENDER;
BKE_color_managed_colorspace_settings_init(&ima->colorspace_settings);
}
return ima;
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
/* Get the ibuf from an image cache by it's index and frame.
* Local use here only.
*
* Returns referenced image buffer if it exists, callee is to
* call IMB_freeImBuf to de-reference the image buffer after
* it's done handling it.
*/
static ImBuf *image_get_cached_ibuf_for_index_frame(Image *ima, int index, int frame)
{
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
if (index != IMA_NO_INDEX) {
index = IMA_MAKE_INDEX(frame, index);
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
return imagecache_get(ima, index);
}
/* no ima->ibuf anymore, but listbase */
static void image_assign_ibuf(Image *ima, ImBuf *ibuf, int index, int frame)
{
if (ibuf) {
if (index != IMA_NO_INDEX)
index = IMA_MAKE_INDEX(frame, index);
2012-06-23 23:22:19 +00:00
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
imagecache_put(ima, index, ibuf);
}
}
2002-10-12 11:37:38 +00:00
/* empty image block, of similar type and filename */
Image *BKE_image_copy(Main *bmain, Image *ima)
{
Image *nima = image_alloc(bmain, ima->id.name + 2, ima->source, ima->type);
BLI_strncpy(nima->name, ima->name, sizeof(ima->name));
nima->flag = ima->flag;
nima->tpageflag = ima->tpageflag;
2012-06-23 23:22:19 +00:00
nima->gen_x = ima->gen_x;
nima->gen_y = ima->gen_y;
nima->gen_type = ima->gen_type;
copy_v4_v4(nima->gen_color, ima->gen_color);
nima->animspeed = ima->animspeed;
nima->aspx = ima->aspx;
nima->aspy = ima->aspy;
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
BKE_color_managed_colorspace_settings_copy(&nima->colorspace_settings, &ima->colorspace_settings);
if (ima->packedfile)
nima->packedfile = dupPackedFile(ima->packedfile);
if (ima->id.lib) {
BKE_id_lib_local_paths(bmain, ima->id.lib, &nima->id);
}
return nima;
}
static void extern_local_image(Image *UNUSED(ima))
{
/* Nothing to do: images don't link to other IDs. This function exists to
* match id_make_local pattern. */
}
void BKE_image_make_local(struct Image *ima)
{
Main *bmain = G.main;
Tex *tex;
Brush *brush;
Mesh *me;
bool is_local = false, is_lib = false;
/* - only lib users: do nothing
* - only local users: set flag
* - mixed: make copy
*/
if (ima->id.lib == NULL) return;
/* Can't take short cut here: must check meshes at least because of bogus
* texface ID refs. - z0r */
#if 0
if (ima->id.us == 1) {
id_clear_lib_data(bmain, &ima->id);
extern_local_image(ima);
return;
}
#endif
for (tex = bmain->tex.first; tex; tex = tex->id.next) {
if (tex->ima == ima) {
if (tex->id.lib) is_lib = true;
else is_local = true;
}
}
for (brush = bmain->brush.first; brush; brush = brush->id.next) {
if (brush->clone.image == ima) {
if (brush->id.lib) is_lib = true;
else is_local = true;
}
}
for (me = bmain->mesh.first; me; me = me->id.next) {
if (me->mtface) {
MTFace *tface;
int a, i;
for (i = 0; i < me->fdata.totlayer; i++) {
if (me->fdata.layers[i].type == CD_MTFACE) {
tface = (MTFace *)me->fdata.layers[i].data;
for (a = 0; a < me->totface; a++, tface++) {
if (tface->tpage == ima) {
if (me->id.lib) is_lib = true;
else is_local = true;
}
}
}
}
}
if (me->mtpoly) {
MTexPoly *mtpoly;
int a, i;
for (i = 0; i < me->pdata.totlayer; i++) {
if (me->pdata.layers[i].type == CD_MTEXPOLY) {
mtpoly = (MTexPoly *)me->pdata.layers[i].data;
for (a = 0; a < me->totpoly; a++, mtpoly++) {
if (mtpoly->tpage == ima) {
if (me->id.lib) is_lib = true;
else is_local = true;
}
}
}
}
}
}
if (is_local && is_lib == false) {
id_clear_lib_data(bmain, &ima->id);
extern_local_image(ima);
}
else if (is_local && is_lib) {
Image *ima_new = BKE_image_copy(bmain, ima);
ima_new->id.us = 0;
/* Remap paths of new ID using old library as base. */
BKE_id_lib_local_paths(bmain, ima->id.lib, &ima_new->id);
tex = bmain->tex.first;
while (tex) {
if (tex->id.lib == NULL) {
if (tex->ima == ima) {
tex->ima = ima_new;
ima_new->id.us++;
ima->id.us--;
}
}
tex = tex->id.next;
}
brush = bmain->brush.first;
while (brush) {
if (brush->id.lib == NULL) {
if (brush->clone.image == ima) {
brush->clone.image = ima_new;
ima_new->id.us++;
ima->id.us--;
}
}
brush = brush->id.next;
}
/* Transfer references in texfaces. Texfaces don't add to image ID
* user count *unless* there are no other users. See
* readfile.c:lib_link_mtface. */
me = bmain->mesh.first;
while (me) {
if (me->mtface) {
MTFace *tface;
int a, i;
for (i = 0; i < me->fdata.totlayer; i++) {
if (me->fdata.layers[i].type == CD_MTFACE) {
tface = (MTFace *)me->fdata.layers[i].data;
for (a = 0; a < me->totface; a++, tface++) {
if (tface->tpage == ima) {
tface->tpage = ima_new;
if (ima_new->id.us == 0) {
tface->tpage->id.us = 1;
}
id_lib_extern((ID *)ima_new);
}
}
}
}
}
if (me->mtpoly) {
MTexPoly *mtpoly;
int a, i;
for (i = 0; i < me->pdata.totlayer; i++) {
if (me->pdata.layers[i].type == CD_MTEXPOLY) {
mtpoly = (MTexPoly *)me->pdata.layers[i].data;
for (a = 0; a < me->totpoly; a++, mtpoly++) {
if (mtpoly->tpage == ima) {
mtpoly->tpage = ima_new;
if (ima_new->id.us == 0) {
mtpoly->tpage->id.us = 1;
}
id_lib_extern((ID *)ima_new);
}
}
}
}
}
me = me->id.next;
}
}
}
void BKE_image_merge(Image *dest, Image *source)
{
/* sanity check */
if (dest && source && dest != source) {
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
BLI_spin_lock(&image_spin);
if (source->cache != NULL) {
struct MovieCacheIter *iter;
iter = IMB_moviecacheIter_new(source->cache);
while (!IMB_moviecacheIter_done(iter)) {
ImBuf *ibuf = IMB_moviecacheIter_getImBuf(iter);
ImageCacheKey *key = IMB_moviecacheIter_getUserKey(iter);
imagecache_put(dest, key->index, ibuf);
IMB_moviecacheIter_step(iter);
}
IMB_moviecacheIter_free(iter);
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
BLI_spin_unlock(&image_spin);
2012-06-23 23:22:19 +00:00
BKE_libblock_free(G.main, source);
}
}
/* note, we could be clever and scale all imbuf's but since some are mipmaps its not so simple */
2014-02-03 18:55:59 +11:00
bool BKE_image_scale(Image *image, int width, int height)
{
ImBuf *ibuf;
void *lock;
ibuf = BKE_image_acquire_ibuf(image, NULL, &lock);
2012-06-29 10:52:37 +00:00
if (ibuf) {
IMB_scaleImBuf(ibuf, width, height);
ibuf->userflags |= IB_BITMAPDIRTY;
}
BKE_image_release_ibuf(image, ibuf, lock);
2012-06-29 10:52:37 +00:00
return (ibuf != NULL);
}
Alpha premul pipeline cleanup This assumptions are now made: - Internally float buffers are always linear alpha-premul colors - Readers should worry about delivering float buffers with that assumptions. - There's an input image setting to say whether it's stored with straight/premul alpha on the disk. - Byte buffers are now assumed have straight alpha, readers should deliver straight alpha. Some implementation details: - Removed scene's color unpremultiply setting, which was very much confusing and was wrong for default settings. Now all renderers assumes to deliver premultiplied alpha. - IMB_buffer_byte_from_float will now linearize alpha when converting from buffer. - Sequencer's effects were changed to assume bytes have got straight alpha. Most of effects will work with bytes still, however for glow it was more tricky to avoid data loss, so there's a commented out glow implementation which converts byte buffer to floats first, operates on floats and returns bytes back. It's slower and not sure if it should actually be used -- who're using glow on alpha anyway? - Sequencer modifiers should also be working nice with straight bytes now. - GLSL preview will predivide float textures to make nice shading, shading with byte textures worked nice (GLSL was assuming straight alpha). - Blender Internal will set alpha=1 to the whole sky. The same happens in Cycles and there's no way to avoid this -- sky is neither straight nor premul and doesn't fit color pipeline well. - Straight alpha mode for render result was also eliminated. - Conversion to correct alpha need to be done before linearizing float buffer. - TIFF will now load and save files with proper alpha mode setting in file meta data header. - Remove Use Alpha from texture mapping and replaced with image datablock setting. Behaves much more predictable and clear from code point of view and solves possible regressions when non-premultiplied images were used as textures with ignoring alpha channel.
2012-12-31 13:52:13 +00:00
static void image_init_color_management(Image *ima)
{
ImBuf *ibuf;
char name[FILE_MAX];
BKE_image_user_file_path(NULL, ima, name);
/* will set input color space to image format default's */
ibuf = IMB_loadiffname(name, IB_test | IB_alphamode_detect, ima->colorspace_settings.name);
if (ibuf) {
if (ibuf->flags & IB_alphamode_premul)
ima->alpha_mode = IMA_ALPHA_PREMUL;
else
ima->alpha_mode = IMA_ALPHA_STRAIGHT;
IMB_freeImBuf(ibuf);
}
}
char BKE_image_alpha_mode_from_extension_ex(const char *filepath)
{
if (BLI_testextensie_n(filepath, ".exr", ".cin", ".dpx", ".hdr", NULL)) {
return IMA_ALPHA_PREMUL;
}
else {
return IMA_ALPHA_STRAIGHT;
}
}
void BKE_image_alpha_mode_from_extension(Image *image)
{
image->alpha_mode = BKE_image_alpha_mode_from_extension_ex(image->name);
}
Image *BKE_image_load(Main *bmain, const char *filepath)
2002-10-12 11:37:38 +00:00
{
Image *ima;
int file, len;
const char *libname;
char str[FILE_MAX];
BLI_strncpy(str, filepath, sizeof(str));
BLI_path_abs(str, bmain->name);
/* exists? */
file = BLI_open(str, O_BINARY | O_RDONLY, 0);
if (file == -1)
return NULL;
2002-10-12 11:37:38 +00:00
close(file);
/* create a short library name */
len = strlen(filepath);
while (len > 0 && filepath[len - 1] != '/' && filepath[len - 1] != '\\') len--;
libname = filepath + len;
ima = image_alloc(bmain, libname, IMA_SRC_FILE, IMA_TYPE_IMAGE);
BLI_strncpy(ima->name, filepath, sizeof(ima->name));
if (BLI_testextensie_array(filepath, imb_ext_movie))
ima->source = IMA_SRC_MOVIE;
Alpha premul pipeline cleanup This assumptions are now made: - Internally float buffers are always linear alpha-premul colors - Readers should worry about delivering float buffers with that assumptions. - There's an input image setting to say whether it's stored with straight/premul alpha on the disk. - Byte buffers are now assumed have straight alpha, readers should deliver straight alpha. Some implementation details: - Removed scene's color unpremultiply setting, which was very much confusing and was wrong for default settings. Now all renderers assumes to deliver premultiplied alpha. - IMB_buffer_byte_from_float will now linearize alpha when converting from buffer. - Sequencer's effects were changed to assume bytes have got straight alpha. Most of effects will work with bytes still, however for glow it was more tricky to avoid data loss, so there's a commented out glow implementation which converts byte buffer to floats first, operates on floats and returns bytes back. It's slower and not sure if it should actually be used -- who're using glow on alpha anyway? - Sequencer modifiers should also be working nice with straight bytes now. - GLSL preview will predivide float textures to make nice shading, shading with byte textures worked nice (GLSL was assuming straight alpha). - Blender Internal will set alpha=1 to the whole sky. The same happens in Cycles and there's no way to avoid this -- sky is neither straight nor premul and doesn't fit color pipeline well. - Straight alpha mode for render result was also eliminated. - Conversion to correct alpha need to be done before linearizing float buffer. - TIFF will now load and save files with proper alpha mode setting in file meta data header. - Remove Use Alpha from texture mapping and replaced with image datablock setting. Behaves much more predictable and clear from code point of view and solves possible regressions when non-premultiplied images were used as textures with ignoring alpha channel.
2012-12-31 13:52:13 +00:00
image_init_color_management(ima);
return ima;
}
/* checks if image was already loaded, then returns same image */
/* otherwise creates new. */
/* does not load ibuf itself */
/* pass on optional frame for #name images */
Image *BKE_image_load_exists_ex(const char *filepath, bool *r_exists)
{
Image *ima;
char str[FILE_MAX], strtest[FILE_MAX];
2012-06-23 23:22:19 +00:00
BLI_strncpy(str, filepath, sizeof(str));
BLI_path_abs(str, G.main->name);
/* first search an identical image */
for (ima = G.main->image.first; ima; ima = ima->id.next) {
if (ima->source != IMA_SRC_VIEWER && ima->source != IMA_SRC_GENERATED) {
BLI_strncpy(strtest, ima->name, sizeof(ima->name));
BLI_path_abs(strtest, ID_BLEND_PATH(G.main, &ima->id));
2012-06-23 23:22:19 +00:00
if (BLI_path_cmp(strtest, str) == 0) {
if (ima->anim == NULL || ima->id.us == 0) {
ima->id.us++; /* officially should not, it doesn't link here! */
if (ima->ok == 0)
ima->ok = IMA_OK;
if (r_exists)
*r_exists = true;
return ima;
}
2002-10-12 11:37:38 +00:00
}
}
}
if (r_exists)
*r_exists = false;
return BKE_image_load(G.main, filepath);
2002-10-12 11:37:38 +00:00
}
Image *BKE_image_load_exists(const char *filepath)
{
return BKE_image_load_exists_ex(filepath, NULL);
}
static ImBuf *add_ibuf_size(unsigned int width, unsigned int height, const char *name, int depth, int floatbuf, short gen_type,
const float color[4], ColorManagedColorspaceSettings *colorspace_settings)
{
ImBuf *ibuf;
unsigned char *rect = NULL;
float *rect_float = NULL;
2012-06-23 23:22:19 +00:00
if (floatbuf) {
ibuf = IMB_allocImBuf(width, height, depth, IB_rectfloat);
if (colorspace_settings->name[0] == '\0') {
const char *colorspace = IMB_colormanagement_role_colorspace_name_get(COLOR_ROLE_DEFAULT_FLOAT);
BLI_strncpy(colorspace_settings->name, colorspace, sizeof(colorspace_settings->name));
}
if (ibuf != NULL) {
rect_float = ibuf->rect_float;
IMB_colormanagement_check_is_data(ibuf, colorspace_settings->name);
}
}
else {
ibuf = IMB_allocImBuf(width, height, depth, IB_rect);
if (colorspace_settings->name[0] == '\0') {
const char *colorspace = IMB_colormanagement_role_colorspace_name_get(COLOR_ROLE_DEFAULT_BYTE);
BLI_strncpy(colorspace_settings->name, colorspace, sizeof(colorspace_settings->name));
}
if (ibuf != NULL) {
rect = (unsigned char *)ibuf->rect;
IMB_colormanagement_assign_rect_colorspace(ibuf, colorspace_settings->name);
}
}
if (!ibuf) {
return NULL;
}
2012-06-23 23:22:19 +00:00
BLI_strncpy(ibuf->name, name, sizeof(ibuf->name));
ibuf->userflags |= IB_BITMAPDIRTY;
2012-06-23 23:22:19 +00:00
switch (gen_type) {
case IMA_GENTYPE_GRID:
BKE_image_buf_fill_checker(rect, rect_float, width, height);
break;
case IMA_GENTYPE_GRID_COLOR:
BKE_image_buf_fill_checker_color(rect, rect_float, width, height);
break;
default:
BKE_image_buf_fill_color(rect, rect_float, width, height, color);
break;
}
if (rect_float) {
/* both byte and float buffers are filling in sRGB space, need to linearize float buffer after BKE_image_buf_fill* functions */
IMB_buffer_float_from_float(rect_float, rect_float, ibuf->channels, IB_PROFILE_LINEAR_RGB, IB_PROFILE_SRGB,
true, ibuf->x, ibuf->y, ibuf->x, ibuf->x);
}
return ibuf;
}
/* adds new image block, creates ImBuf and initializes color */
Image *BKE_image_add_generated(Main *bmain, unsigned int width, unsigned int height, const char *name, int depth, int floatbuf, short gen_type, const float color[4])
{
/* on save, type is changed to FILE in editsima.c */
Image *ima = image_alloc(bmain, name, IMA_SRC_GENERATED, IMA_TYPE_UV_TEST);
2012-06-23 23:22:19 +00:00
if (ima) {
ImBuf *ibuf;
2012-06-23 23:22:19 +00:00
2012-03-18 07:38:51 +00:00
/* BLI_strncpy(ima->name, name, FILE_MAX); */ /* don't do this, this writes in ain invalid filepath! */
ima->gen_x = width;
ima->gen_y = height;
ima->gen_type = gen_type;
ima->gen_flag |= (floatbuf ? IMA_GEN_FLOAT : 0);
ima->gen_depth = depth;
copy_v4_v4(ima->gen_color, color);
2012-06-23 23:22:19 +00:00
ibuf = add_ibuf_size(width, height, ima->name, depth, floatbuf, gen_type, color, &ima->colorspace_settings);
image_assign_ibuf(ima, ibuf, IMA_NO_INDEX, 0);
2012-06-23 23:22:19 +00:00
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
/* image_assign_ibuf puts buffer to the cache, which increments user counter. */
IMB_freeImBuf(ibuf);
ima->ok = IMA_OK_LOADED;
}
return ima;
}
/* Create an image image from ibuf. The refcount of ibuf is increased,
* caller should take care to drop its reference by calling
* IMB_freeImBuf if needed. */
Image *BKE_image_add_from_imbuf(ImBuf *ibuf)
{
/* on save, type is changed to FILE in editsima.c */
Image *ima;
ima = image_alloc(G.main, BLI_path_basename(ibuf->name), IMA_SRC_FILE, IMA_TYPE_IMAGE);
if (ima) {
BLI_strncpy(ima->name, ibuf->name, FILE_MAX);
image_assign_ibuf(ima, ibuf, IMA_NO_INDEX, 0);
ima->ok = IMA_OK_LOADED;
}
return ima;
}
/* packs rect from memory as PNG */
void BKE_image_memorypack(Image *ima)
{
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ImBuf *ibuf = image_get_cached_ibuf_for_index_frame(ima, IMA_NO_INDEX, 0);
2012-06-23 23:22:19 +00:00
if (ibuf == NULL)
return;
if (ima->packedfile) {
freePackedFile(ima->packedfile);
ima->packedfile = NULL;
}
2012-06-23 23:22:19 +00:00
ibuf->ftype = PNG;
ibuf->planes = R_IMF_PLANES_RGBA;
2012-06-23 23:22:19 +00:00
IMB_saveiff(ibuf, ibuf->name, IB_rect | IB_mem);
if (ibuf->encodedbuffer == NULL) {
printf("memory save for pack error\n");
}
else {
PackedFile *pf = MEM_callocN(sizeof(*pf), "PackedFile");
2012-06-23 23:22:19 +00:00
pf->data = ibuf->encodedbuffer;
pf->size = ibuf->encodedsize;
ima->packedfile = pf;
ibuf->encodedbuffer = NULL;
ibuf->encodedsize = 0;
ibuf->userflags &= ~IB_BITMAPDIRTY;
2012-06-23 23:22:19 +00:00
if (ima->source == IMA_SRC_GENERATED) {
ima->source = IMA_SRC_FILE;
ima->type = IMA_TYPE_IMAGE;
}
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
IMB_freeImBuf(ibuf);
}
void BKE_image_tag_time(Image *ima)
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
{
ima->lastused = PIL_check_seconds_timer_i();
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
}
#if 0
2012-06-23 23:22:19 +00:00
static void tag_all_images_time()
{
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
Image *ima;
int ctime = PIL_check_seconds_timer_i();
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
ima = G.main->image.first;
while (ima) {
if (ima->bindcode || ima->repbind || ima->ibufs.first) {
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
ima->lastused = ctime;
}
}
}
#endif
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
static uintptr_t image_mem_size(Image *image)
{
uintptr_t size = 0;
/* viewers have memory depending on other rules, has no valid rect pointer */
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
if (image->source == IMA_SRC_VIEWER)
return 0;
2012-06-23 23:22:19 +00:00
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
BLI_spin_lock(&image_spin);
if (image->cache != NULL) {
struct MovieCacheIter *iter = IMB_moviecacheIter_new(image->cache);
while (!IMB_moviecacheIter_done(iter)) {
ImBuf *ibuf = IMB_moviecacheIter_getImBuf(iter);
ImBuf *ibufm;
int level;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
if (ibuf->rect) {
size += MEM_allocN_len(ibuf->rect);
}
if (ibuf->rect_float) {
size += MEM_allocN_len(ibuf->rect_float);
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
for (level = 0; level < IB_MIPMAP_LEVELS; level++) {
ibufm = ibuf->mipmap[level];
if (ibufm) {
if (ibufm->rect) {
size += MEM_allocN_len(ibufm->rect);
}
if (ibufm->rect_float) {
size += MEM_allocN_len(ibufm->rect_float);
}
}
}
IMB_moviecacheIter_step(iter);
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
IMB_moviecacheIter_free(iter);
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
BLI_spin_unlock(&image_spin);
return size;
}
void BKE_image_print_memlist(void)
{
Image *ima;
uintptr_t size, totsize = 0;
for (ima = G.main->image.first; ima; ima = ima->id.next)
totsize += image_mem_size(ima);
printf("\ntotal image memory len: %.3f MB\n", (double)totsize / (double)(1024 * 1024));
for (ima = G.main->image.first; ima; ima = ima->id.next) {
size = image_mem_size(ima);
if (size)
printf("%s len: %.3f MB\n", ima->id.name + 2, (double)size / (double)(1024 * 1024));
}
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
static bool imagecache_check_dirty(ImBuf *ibuf, void *UNUSED(userkey), void *UNUSED(userdata))
{
return (ibuf->userflags & IB_BITMAPDIRTY) == 0;
}
void BKE_image_free_all_textures(void)
{
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
#undef CHECK_FREED_SIZE
Tex *tex;
Image *ima;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
#ifdef CHECK_FREED_SIZE
uintptr_t tot_freed_size = 0;
#endif
2012-06-23 23:22:19 +00:00
for (ima = G.main->image.first; ima; ima = ima->id.next)
ima->id.flag &= ~LIB_DOIT;
2012-06-23 23:22:19 +00:00
for (tex = G.main->tex.first; tex; tex = tex->id.next)
if (tex->ima)
tex->ima->id.flag |= LIB_DOIT;
2012-06-23 23:22:19 +00:00
for (ima = G.main->image.first; ima; ima = ima->id.next) {
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
if (ima->cache && (ima->id.flag & LIB_DOIT)) {
#ifdef CHECK_FREED_SIZE
uintptr_t old_size = image_mem_size(ima);
#endif
2012-06-23 23:22:19 +00:00
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
IMB_moviecache_cleanup(ima->cache, imagecache_check_dirty, NULL);
2012-06-23 23:22:19 +00:00
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
#ifdef CHECK_FREED_SIZE
tot_freed_size += old_size - image_mem_size(ima);
#endif
}
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
#ifdef CHECK_FREED_SIZE
printf("%s: freed total %lu MB\n", __func__, tot_freed_size / (1024 * 1024));
#endif
}
static bool imagecache_check_free_anim(ImBuf *ibuf, void *UNUSED(userkey), void *userdata)
{
int except_frame = *(int *)userdata;
return (ibuf->userflags & IB_BITMAPDIRTY) == 0 &&
(ibuf->index != IMA_NO_INDEX) &&
(except_frame != IMA_INDEX_FRAME(ibuf->index));
}
2002-10-12 11:37:38 +00:00
/* except_frame is weak, only works for seqs without offset... */
void BKE_image_free_anim_ibufs(Image *ima, int except_frame)
{
BLI_spin_lock(&image_spin);
if (ima->cache != NULL) {
IMB_moviecache_cleanup(ima->cache, imagecache_check_free_anim, &except_frame);
}
BLI_spin_unlock(&image_spin);
}
void BKE_image_all_free_anim_ibufs(int cfra)
{
Image *ima;
2012-06-23 23:22:19 +00:00
for (ima = G.main->image.first; ima; ima = ima->id.next)
if (BKE_image_is_animated(ima))
BKE_image_free_anim_ibufs(ima, cfra);
}
/* *********** READ AND WRITE ************** */
2002-10-12 11:37:38 +00:00
int BKE_image_imtype_to_ftype(const char imtype)
{
if (imtype == R_IMF_IMTYPE_TARGA)
return TGA;
else if (imtype == R_IMF_IMTYPE_RAWTGA)
return RAWTGA;
else if (imtype == R_IMF_IMTYPE_IRIS)
return IMAGIC;
#ifdef WITH_HDR
else if (imtype == R_IMF_IMTYPE_RADHDR)
return RADHDR;
#endif
else if (imtype == R_IMF_IMTYPE_PNG)
return PNG | 15;
#ifdef WITH_DDS
else if (imtype == R_IMF_IMTYPE_DDS)
return DDS;
#endif
else if (imtype == R_IMF_IMTYPE_BMP)
return BMP;
#ifdef WITH_TIFF
else if (imtype == R_IMF_IMTYPE_TIFF)
return TIF;
#endif
else if (imtype == R_IMF_IMTYPE_OPENEXR || imtype == R_IMF_IMTYPE_MULTILAYER)
return OPENEXR;
#ifdef WITH_CINEON
else if (imtype == R_IMF_IMTYPE_CINEON)
return CINEON;
else if (imtype == R_IMF_IMTYPE_DPX)
return DPX;
#endif
#ifdef WITH_OPENJPEG
else if (imtype == R_IMF_IMTYPE_JP2)
return JP2;
#endif
else
return JPG | 90;
}
char BKE_image_ftype_to_imtype(const int ftype)
{
if (ftype == 0)
return R_IMF_IMTYPE_TARGA;
2012-06-23 23:22:19 +00:00
else if (ftype == IMAGIC)
return R_IMF_IMTYPE_IRIS;
#ifdef WITH_HDR
else if (ftype & RADHDR)
return R_IMF_IMTYPE_RADHDR;
#endif
else if (ftype & PNG)
return R_IMF_IMTYPE_PNG;
#ifdef WITH_DDS
else if (ftype & DDS)
return R_IMF_IMTYPE_DDS;
#endif
else if (ftype & BMP)
return R_IMF_IMTYPE_BMP;
#ifdef WITH_TIFF
else if (ftype & TIF)
return R_IMF_IMTYPE_TIFF;
#endif
else if (ftype & OPENEXR)
return R_IMF_IMTYPE_OPENEXR;
#ifdef WITH_CINEON
else if (ftype & CINEON)
return R_IMF_IMTYPE_CINEON;
else if (ftype & DPX)
return R_IMF_IMTYPE_DPX;
#endif
else if (ftype & TGA)
return R_IMF_IMTYPE_TARGA;
else if (ftype & RAWTGA)
return R_IMF_IMTYPE_RAWTGA;
#ifdef WITH_OPENJPEG
else if (ftype & JP2)
return R_IMF_IMTYPE_JP2;
#endif
else
return R_IMF_IMTYPE_JPEG90;
}
2013-03-09 05:35:49 +00:00
bool BKE_imtype_is_movie(const char imtype)
2002-10-12 11:37:38 +00:00
{
switch (imtype) {
case R_IMF_IMTYPE_AVIRAW:
case R_IMF_IMTYPE_AVIJPEG:
case R_IMF_IMTYPE_QUICKTIME:
case R_IMF_IMTYPE_FFMPEG:
case R_IMF_IMTYPE_H264:
case R_IMF_IMTYPE_THEORA:
case R_IMF_IMTYPE_XVID:
case R_IMF_IMTYPE_FRAMESERVER:
2013-03-09 05:35:49 +00:00
return true;
2002-10-12 11:37:38 +00:00
}
2013-03-09 05:35:49 +00:00
return false;
2002-10-12 11:37:38 +00:00
}
int BKE_imtype_supports_zbuf(const char imtype)
{
switch (imtype) {
case R_IMF_IMTYPE_IRIZ:
case R_IMF_IMTYPE_OPENEXR: /* but not R_IMF_IMTYPE_MULTILAYER */
return 1;
}
return 0;
}
int BKE_imtype_supports_compress(const char imtype)
{
switch (imtype) {
case R_IMF_IMTYPE_PNG:
return 1;
}
return 0;
}
int BKE_imtype_supports_quality(const char imtype)
{
switch (imtype) {
case R_IMF_IMTYPE_JPEG90:
case R_IMF_IMTYPE_JP2:
case R_IMF_IMTYPE_AVIJPEG:
return 1;
}
return 0;
}
int BKE_imtype_requires_linear_float(const char imtype)
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
{
switch (imtype) {
case R_IMF_IMTYPE_CINEON:
case R_IMF_IMTYPE_DPX:
case R_IMF_IMTYPE_RADHDR:
case R_IMF_IMTYPE_OPENEXR:
case R_IMF_IMTYPE_MULTILAYER:
return true;
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
}
return 0;
}
char BKE_imtype_valid_channels(const char imtype, bool write_file)
{
char chan_flag = IMA_CHAN_FLAG_RGB; /* assume all support rgb */
/* alpha */
switch (imtype) {
case R_IMF_IMTYPE_BMP:
if (write_file) break;
/* fall-through */
case R_IMF_IMTYPE_TARGA:
case R_IMF_IMTYPE_IRIS:
case R_IMF_IMTYPE_PNG:
case R_IMF_IMTYPE_RADHDR:
case R_IMF_IMTYPE_TIFF:
case R_IMF_IMTYPE_OPENEXR:
case R_IMF_IMTYPE_MULTILAYER:
case R_IMF_IMTYPE_DDS:
case R_IMF_IMTYPE_JP2:
case R_IMF_IMTYPE_QUICKTIME:
case R_IMF_IMTYPE_DPX:
chan_flag |= IMA_CHAN_FLAG_ALPHA;
break;
}
/* bw */
switch (imtype) {
case R_IMF_IMTYPE_PNG:
case R_IMF_IMTYPE_JPEG90:
case R_IMF_IMTYPE_TARGA:
case R_IMF_IMTYPE_RAWTGA:
case R_IMF_IMTYPE_TIFF:
case R_IMF_IMTYPE_IRIS:
chan_flag |= IMA_CHAN_FLAG_BW;
break;
}
return chan_flag;
}
char BKE_imtype_valid_depths(const char imtype)
{
switch (imtype) {
case R_IMF_IMTYPE_RADHDR:
return R_IMF_CHAN_DEPTH_32;
case R_IMF_IMTYPE_TIFF:
return R_IMF_CHAN_DEPTH_8 | R_IMF_CHAN_DEPTH_16;
case R_IMF_IMTYPE_OPENEXR:
return R_IMF_CHAN_DEPTH_16 | R_IMF_CHAN_DEPTH_32;
case R_IMF_IMTYPE_MULTILAYER:
return R_IMF_CHAN_DEPTH_32;
/* eeh, cineon does some strange 10bits per channel */
case R_IMF_IMTYPE_DPX:
return R_IMF_CHAN_DEPTH_8 | R_IMF_CHAN_DEPTH_10 | R_IMF_CHAN_DEPTH_12 | R_IMF_CHAN_DEPTH_16;
case R_IMF_IMTYPE_CINEON:
return R_IMF_CHAN_DEPTH_10;
case R_IMF_IMTYPE_JP2:
return R_IMF_CHAN_DEPTH_8 | R_IMF_CHAN_DEPTH_12 | R_IMF_CHAN_DEPTH_16;
case R_IMF_IMTYPE_PNG:
return R_IMF_CHAN_DEPTH_8 | R_IMF_CHAN_DEPTH_16;
/* most formats are 8bit only */
default:
return R_IMF_CHAN_DEPTH_8;
}
}
/* string is from command line --render-format arg, keep in sync with
* creator.c help info */
char BKE_imtype_from_arg(const char *imtype_arg)
{
if (STREQ(imtype_arg, "TGA")) return R_IMF_IMTYPE_TARGA;
else if (STREQ(imtype_arg, "IRIS")) return R_IMF_IMTYPE_IRIS;
#ifdef WITH_DDS
else if (STREQ(imtype_arg, "DDS")) return R_IMF_IMTYPE_DDS;
#endif
else if (STREQ(imtype_arg, "JPEG")) return R_IMF_IMTYPE_JPEG90;
else if (STREQ(imtype_arg, "IRIZ")) return R_IMF_IMTYPE_IRIZ;
else if (STREQ(imtype_arg, "RAWTGA")) return R_IMF_IMTYPE_RAWTGA;
else if (STREQ(imtype_arg, "AVIRAW")) return R_IMF_IMTYPE_AVIRAW;
else if (STREQ(imtype_arg, "AVIJPEG")) return R_IMF_IMTYPE_AVIJPEG;
else if (STREQ(imtype_arg, "PNG")) return R_IMF_IMTYPE_PNG;
else if (STREQ(imtype_arg, "QUICKTIME")) return R_IMF_IMTYPE_QUICKTIME;
else if (STREQ(imtype_arg, "BMP")) return R_IMF_IMTYPE_BMP;
#ifdef WITH_HDR
else if (STREQ(imtype_arg, "HDR")) return R_IMF_IMTYPE_RADHDR;
#endif
#ifdef WITH_TIFF
else if (STREQ(imtype_arg, "TIFF")) return R_IMF_IMTYPE_TIFF;
#endif
#ifdef WITH_OPENEXR
else if (STREQ(imtype_arg, "EXR")) return R_IMF_IMTYPE_OPENEXR;
else if (STREQ(imtype_arg, "MULTILAYER")) return R_IMF_IMTYPE_MULTILAYER;
#endif
else if (STREQ(imtype_arg, "MPEG")) return R_IMF_IMTYPE_FFMPEG;
else if (STREQ(imtype_arg, "FRAMESERVER")) return R_IMF_IMTYPE_FRAMESERVER;
#ifdef WITH_CINEON
else if (STREQ(imtype_arg, "CINEON")) return R_IMF_IMTYPE_CINEON;
else if (STREQ(imtype_arg, "DPX")) return R_IMF_IMTYPE_DPX;
#endif
#ifdef WITH_OPENJPEG
else if (STREQ(imtype_arg, "JP2")) return R_IMF_IMTYPE_JP2;
#endif
else return R_IMF_IMTYPE_INVALID;
}
static bool image_path_ensure_ext(char *string, const char imtype, const ImageFormatData *im_format)
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
{
const char *extension = NULL;
const char *extension_test;
2012-12-23 16:09:26 +00:00
(void)im_format; /* may be unused, depends on build options */
2012-06-23 23:22:19 +00:00
if (imtype == R_IMF_IMTYPE_IRIS) {
if (!BLI_testextensie(string, extension_test = ".rgb"))
extension = extension_test;
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
}
else if (imtype == R_IMF_IMTYPE_IRIZ) {
if (!BLI_testextensie(string, extension_test = ".rgb"))
extension = extension_test;
}
#ifdef WITH_HDR
else if (imtype == R_IMF_IMTYPE_RADHDR) {
if (!BLI_testextensie(string, extension_test = ".hdr"))
extension = extension_test;
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
}
#endif
else if (ELEM(imtype, R_IMF_IMTYPE_PNG, R_IMF_IMTYPE_FFMPEG, R_IMF_IMTYPE_H264, R_IMF_IMTYPE_THEORA, R_IMF_IMTYPE_XVID)) {
if (!BLI_testextensie(string, extension_test = ".png"))
extension = extension_test;
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
}
#ifdef WITH_DDS
else if (imtype == R_IMF_IMTYPE_DDS) {
if (!BLI_testextensie(string, extension_test = ".dds"))
extension = extension_test;
}
#endif
else if (ELEM(imtype, R_IMF_IMTYPE_TARGA, R_IMF_IMTYPE_RAWTGA)) {
if (!BLI_testextensie(string, extension_test = ".tga"))
extension = extension_test;
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
}
else if (imtype == R_IMF_IMTYPE_BMP) {
if (!BLI_testextensie(string, extension_test = ".bmp"))
extension = extension_test;
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
}
#ifdef WITH_TIFF
else if (imtype == R_IMF_IMTYPE_TIFF) {
if (!BLI_testextensie_n(string, extension_test = ".tif", ".tiff", NULL)) {
extension = extension_test;
}
}
#endif
#ifdef WITH_OPENIMAGEIO
else if (imtype == R_IMF_IMTYPE_PSD) {
if (!BLI_testextensie(string, extension_test = ".psd"))
extension = extension_test;
}
#endif
#ifdef WITH_OPENEXR
2012-04-29 15:47:02 +00:00
else if (ELEM(imtype, R_IMF_IMTYPE_OPENEXR, R_IMF_IMTYPE_MULTILAYER)) {
if (!BLI_testextensie(string, extension_test = ".exr"))
extension = extension_test;
}
#endif
#ifdef WITH_CINEON
else if (imtype == R_IMF_IMTYPE_CINEON) {
if (!BLI_testextensie(string, extension_test = ".cin"))
extension = extension_test;
}
else if (imtype == R_IMF_IMTYPE_DPX) {
if (!BLI_testextensie(string, extension_test = ".dpx"))
extension = extension_test;
}
#endif
#ifdef WITH_OPENJPEG
else if (imtype == R_IMF_IMTYPE_JP2) {
if (im_format) {
if (im_format->jp2_codec == R_IMF_JP2_CODEC_JP2) {
if (!BLI_testextensie(string, extension_test = ".jp2"))
extension = extension_test;
}
else if (im_format->jp2_codec == R_IMF_JP2_CODEC_J2K) {
if (!BLI_testextensie(string, extension_test = ".j2c"))
extension = extension_test;
}
else
BLI_assert(!"Unsupported jp2 codec was specified in im_format->jp2_codec");
}
else {
if (!BLI_testextensie(string, extension_test = ".jp2"))
extension = extension_test;
}
}
#endif
else { // R_IMF_IMTYPE_AVIRAW, R_IMF_IMTYPE_AVIJPEG, R_IMF_IMTYPE_JPEG90, R_IMF_IMTYPE_QUICKTIME etc
if (!(BLI_testextensie_n(string, extension_test = ".jpg", ".jpeg", NULL)))
extension = extension_test;
}
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
if (extension) {
/* prefer this in many cases to avoid .png.tga, but in certain cases it breaks */
/* remove any other known image extension */
if (BLI_testextensie_array(string, imb_ext_image) ||
(G.have_quicktime && BLI_testextensie_array(string, imb_ext_image_qt)))
{
return BLI_replace_extension(string, FILE_MAX, extension);
}
else {
return BLI_ensure_extension(string, FILE_MAX, extension);
}
2012-06-23 23:22:19 +00:00
}
else {
return false;
}
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
}
bool BKE_image_path_ensure_ext_from_imformat(char *string, const ImageFormatData *im_format)
{
return image_path_ensure_ext(string, im_format->imtype, im_format);
}
bool BKE_image_path_ensure_ext_from_imtype(char *string, const char imtype)
{
return image_path_ensure_ext(string, imtype, NULL);
}
void BKE_imformat_defaults(ImageFormatData *im_format)
{
memset(im_format, 0, sizeof(*im_format));
im_format->planes = R_IMF_PLANES_RGBA;
im_format->imtype = R_IMF_IMTYPE_PNG;
im_format->depth = R_IMF_CHAN_DEPTH_8;
im_format->quality = 90;
im_format->compress = 15;
BKE_color_managed_display_settings_init(&im_format->display_settings);
BKE_color_managed_view_settings_init(&im_format->view_settings);
}
void BKE_imbuf_to_image_format(struct ImageFormatData *im_format, const ImBuf *imbuf)
{
int ftype = imbuf->ftype & ~IB_CUSTOM_FLAGS_MASK;
int custom_flags = imbuf->ftype & IB_CUSTOM_FLAGS_MASK;
BKE_imformat_defaults(im_format);
/* file type */
if (ftype == IMAGIC)
im_format->imtype = R_IMF_IMTYPE_IRIS;
#ifdef WITH_HDR
else if (ftype == RADHDR)
im_format->imtype = R_IMF_IMTYPE_RADHDR;
#endif
else if (ftype == PNG) {
im_format->imtype = R_IMF_IMTYPE_PNG;
if (custom_flags & PNG_16BIT)
im_format->depth = R_IMF_CHAN_DEPTH_16;
}
#ifdef WITH_DDS
else if (ftype == DDS)
im_format->imtype = R_IMF_IMTYPE_DDS;
#endif
else if (ftype == BMP)
im_format->imtype = R_IMF_IMTYPE_BMP;
#ifdef WITH_TIFF
else if (ftype == TIF) {
im_format->imtype = R_IMF_IMTYPE_TIFF;
if (custom_flags & TIF_16BIT)
im_format->depth = R_IMF_CHAN_DEPTH_16;
}
#endif
#ifdef WITH_OPENEXR
else if (ftype == OPENEXR) {
2012-06-23 23:22:19 +00:00
im_format->imtype = R_IMF_IMTYPE_OPENEXR;
if (custom_flags & OPENEXR_HALF)
im_format->depth = R_IMF_CHAN_DEPTH_16;
if (custom_flags & OPENEXR_COMPRESS)
2012-06-23 23:22:19 +00:00
im_format->exr_codec = R_IMF_EXR_CODEC_ZIP; // Can't determine compression
if (imbuf->zbuf_float)
im_format->flag |= R_IMF_FLAG_ZBUF;
}
#endif
#ifdef WITH_CINEON
else if (ftype == CINEON)
im_format->imtype = R_IMF_IMTYPE_CINEON;
else if (ftype == DPX)
im_format->imtype = R_IMF_IMTYPE_DPX;
#endif
else if (ftype == TGA) {
im_format->imtype = R_IMF_IMTYPE_TARGA;
}
else if (ftype == RAWTGA) {
im_format->imtype = R_IMF_IMTYPE_RAWTGA;
}
#ifdef WITH_OPENJPEG
else if (ftype & JP2) {
im_format->imtype = R_IMF_IMTYPE_JP2;
im_format->quality = custom_flags & ~JPG_MSK;
if (ftype & JP2_16BIT)
im_format->depth = R_IMF_CHAN_DEPTH_16;
else if (ftype & JP2_12BIT)
im_format->depth = R_IMF_CHAN_DEPTH_12;
if (ftype & JP2_YCC)
im_format->jp2_flag |= R_IMF_JP2_FLAG_YCC;
if (ftype & JP2_CINE) {
im_format->jp2_flag |= R_IMF_JP2_FLAG_CINE_PRESET;
if (ftype & JP2_CINE_48FPS)
im_format->jp2_flag |= R_IMF_JP2_FLAG_CINE_48;
}
if (ftype & JP2_JP2)
im_format->jp2_codec = R_IMF_JP2_CODEC_JP2;
else if (ftype & JP2_J2K)
im_format->jp2_codec = R_IMF_JP2_CODEC_J2K;
else
BLI_assert(!"Unsupported jp2 codec was specified in file type");
}
#endif
else {
im_format->imtype = R_IMF_IMTYPE_JPEG90;
im_format->quality = custom_flags & ~JPG_MSK;
}
/* planes */
/* TODO(sergey): Channels doesn't correspond actual planes used for image buffer
* For example byte buffer will have 4 channels but it might easily
* be BW or RGB image.
*
* Need to use im_format->planes = imbuf->planes instead?
*/
switch (imbuf->channels) {
case 0:
case 4: im_format->planes = R_IMF_PLANES_RGBA;
2012-06-23 23:22:19 +00:00
break;
case 3: im_format->planes = R_IMF_PLANES_RGB;
2012-06-23 23:22:19 +00:00
break;
case 1: im_format->planes = R_IMF_PLANES_BW;
2012-06-23 23:22:19 +00:00
break;
default: im_format->planes = R_IMF_PLANES_RGB;
break;
}
}
#define STAMP_NAME_SIZE ((MAX_ID_NAME - 2) + 16)
/* could allow access externally - 512 is for long names,
* STAMP_NAME_SIZE is for id names, allowing them some room for description */
typedef struct StampData {
char file[512];
char note[512];
char date[512];
char marker[512];
char time[512];
char frame[512];
char camera[STAMP_NAME_SIZE];
char cameralens[STAMP_NAME_SIZE];
char scene[STAMP_NAME_SIZE];
char strip[STAMP_NAME_SIZE];
char rendertime[STAMP_NAME_SIZE];
} StampData;
#undef STAMP_NAME_SIZE
static void stampdata(Scene *scene, Object *camera, StampData *stamp_data, int do_prefix)
{
char text[256];
struct tm *tl;
time_t t;
if (scene->r.stamp & R_STAMP_FILENAME) {
BLI_snprintf(stamp_data->file, sizeof(stamp_data->file), do_prefix ? "File %s" : "%s", G.relbase_valid ? G.main->name : "<untitled>");
}
else {
stamp_data->file[0] = '\0';
}
2012-06-23 23:22:19 +00:00
if (scene->r.stamp & R_STAMP_NOTE) {
/* Never do prefix for Note */
BLI_snprintf(stamp_data->note, sizeof(stamp_data->note), "%s", scene->r.stamp_udata);
}
else {
stamp_data->note[0] = '\0';
}
2012-06-23 23:22:19 +00:00
if (scene->r.stamp & R_STAMP_DATE) {
t = time(NULL);
tl = localtime(&t);
BLI_snprintf(text, sizeof(text), "%04d/%02d/%02d %02d:%02d:%02d", tl->tm_year + 1900, tl->tm_mon + 1, tl->tm_mday, tl->tm_hour, tl->tm_min, tl->tm_sec);
BLI_snprintf(stamp_data->date, sizeof(stamp_data->date), do_prefix ? "Date %s" : "%s", text);
}
else {
stamp_data->date[0] = '\0';
}
2012-06-23 23:22:19 +00:00
if (scene->r.stamp & R_STAMP_MARKER) {
const char *name = BKE_scene_find_last_marker_name(scene, CFRA);
if (name) BLI_strncpy(text, name, sizeof(text));
else BLI_strncpy(text, "<none>", sizeof(text));
BLI_snprintf(stamp_data->marker, sizeof(stamp_data->marker), do_prefix ? "Marker %s" : "%s", text);
}
else {
stamp_data->marker[0] = '\0';
}
2012-06-23 23:22:19 +00:00
if (scene->r.stamp & R_STAMP_TIME) {
const short timecode_style = USER_TIMECODE_SMPTE_FULL;
BLI_timecode_string_from_time(text, sizeof(text), 0, FRA2TIME(scene->r.cfra), FPS, timecode_style);
BLI_snprintf(stamp_data->time, sizeof(stamp_data->time), do_prefix ? "Timecode %s" : "%s", text);
}
else {
stamp_data->time[0] = '\0';
}
2012-06-23 23:22:19 +00:00
if (scene->r.stamp & R_STAMP_FRAME) {
char fmtstr[32];
int digits = 1;
2012-06-23 23:22:19 +00:00
if (scene->r.efra > 9)
digits = 1 + (int) log10(scene->r.efra);
BLI_snprintf(fmtstr, sizeof(fmtstr), do_prefix ? "Frame %%0%di" : "%%0%di", digits);
BLI_snprintf(stamp_data->frame, sizeof(stamp_data->frame), fmtstr, scene->r.cfra);
}
else {
stamp_data->frame[0] = '\0';
}
if (scene->r.stamp & R_STAMP_CAMERA) {
BLI_snprintf(stamp_data->camera, sizeof(stamp_data->camera), do_prefix ? "Camera %s" : "%s", camera ? camera->id.name + 2 : "<none>");
}
else {
stamp_data->camera[0] = '\0';
}
if (scene->r.stamp & R_STAMP_CAMERALENS) {
if (camera && camera->type == OB_CAMERA) {
BLI_snprintf(text, sizeof(text), "%.2f", ((Camera *)camera->data)->lens);
}
else {
BLI_strncpy(text, "<none>", sizeof(text));
}
BLI_snprintf(stamp_data->cameralens, sizeof(stamp_data->cameralens), do_prefix ? "Lens %s" : "%s", text);
}
else {
stamp_data->cameralens[0] = '\0';
}
if (scene->r.stamp & R_STAMP_SCENE) {
BLI_snprintf(stamp_data->scene, sizeof(stamp_data->scene), do_prefix ? "Scene %s" : "%s", scene->id.name + 2);
}
else {
stamp_data->scene[0] = '\0';
}
2012-06-23 23:22:19 +00:00
if (scene->r.stamp & R_STAMP_SEQSTRIP) {
Sequence *seq = BKE_sequencer_foreground_frame_get(scene, scene->r.cfra);
2012-06-23 23:22:19 +00:00
if (seq) BLI_strncpy(text, seq->name + 2, sizeof(text));
else BLI_strncpy(text, "<none>", sizeof(text));
BLI_snprintf(stamp_data->strip, sizeof(stamp_data->strip), do_prefix ? "Strip %s" : "%s", text);
}
else {
stamp_data->strip[0] = '\0';
}
{
Render *re = RE_GetRender(scene->id.name);
RenderStats *stats = re ? RE_GetStats(re) : NULL;
if (stats && (scene->r.stamp & R_STAMP_RENDERTIME)) {
BLI_timestr(stats->lastframetime, text, sizeof(text));
BLI_snprintf(stamp_data->rendertime, sizeof(stamp_data->rendertime), do_prefix ? "RenderTime %s" : "%s", text);
}
else {
stamp_data->rendertime[0] = '\0';
}
}
}
void BKE_image_stamp_buf(
Scene *scene, Object *camera,
unsigned char *rect, float *rectf, int width, int height, int channels)
{
struct StampData stamp_data;
float w, h, pad;
int x, y, y_ofs;
float h_fixed;
const int mono = blf_mono_font_render; // XXX
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
struct ColorManagedDisplay *display;
const char *display_device;
/* this could be an argument if we want to operate on non linear float imbuf's
* for now though this is only used for renders which use scene settings */
#define TEXT_SIZE_CHECK(str, w, h) \
((str[0]) && ((void)(h = h_fixed), (w = BLF_width(mono, str, sizeof(str)))))
#define BUFF_MARGIN_X 2
#define BUFF_MARGIN_Y 1
if (!rect && !rectf)
return;
2012-06-23 23:22:19 +00:00
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
display_device = scene->display_settings.display_device;
display = IMB_colormanagement_display_get_named(display_device);
stampdata(scene, camera, &stamp_data, 1);
/* TODO, do_versions */
if (scene->r.stamp_font_id < 8)
scene->r.stamp_font_id = 12;
/* set before return */
BLF_size(mono, scene->r.stamp_font_id, 72);
2012-06-23 23:22:19 +00:00
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
BLF_buffer(mono, rectf, rect, width, height, channels, display);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_buffer_col(mono, scene->r.fg_stamp[0], scene->r.fg_stamp[1], scene->r.fg_stamp[2], 1.0);
pad = BLF_width_max(mono);
/* use 'h_fixed' rather than 'h', aligns better */
h_fixed = BLF_height_max(mono);
y_ofs = -BLF_descender(mono);
x = 0;
y = height;
if (TEXT_SIZE_CHECK(stamp_data.file, w, h)) {
/* Top left corner */
y -= h;
/* also a little of space to the background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
x - BUFF_MARGIN_X, y - BUFF_MARGIN_Y, w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
/* and draw the text. */
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.file);
/* the extra pixel for background. */
y -= BUFF_MARGIN_Y * 2;
}
/* Top left corner, below File */
if (TEXT_SIZE_CHECK(stamp_data.note, w, h)) {
y -= h;
/* and space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
0, y - BUFF_MARGIN_Y, w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.note);
/* the extra pixel for background. */
y -= BUFF_MARGIN_Y * 2;
}
2012-06-23 23:22:19 +00:00
/* Top left corner, below File (or Note) */
if (TEXT_SIZE_CHECK(stamp_data.date, w, h)) {
y -= h;
/* and space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
0, y - BUFF_MARGIN_Y, w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.date);
/* the extra pixel for background. */
y -= BUFF_MARGIN_Y * 2;
}
/* Top left corner, below File, Date or Note */
if (TEXT_SIZE_CHECK(stamp_data.rendertime, w, h)) {
y -= h;
/* and space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
0, y - BUFF_MARGIN_Y, w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.rendertime);
}
x = 0;
y = 0;
/* Bottom left corner, leaving space for timing */
if (TEXT_SIZE_CHECK(stamp_data.marker, w, h)) {
/* extra space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
x - BUFF_MARGIN_X, y - BUFF_MARGIN_Y, w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
/* and pad the text. */
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.marker);
/* space width. */
x += w + pad;
}
2012-06-23 23:22:19 +00:00
/* Left bottom corner */
if (TEXT_SIZE_CHECK(stamp_data.time, w, h)) {
/* extra space for background */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
x - BUFF_MARGIN_X, y, x + w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
/* and pad the text. */
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.time);
/* space width. */
x += w + pad;
}
2012-06-23 23:22:19 +00:00
if (TEXT_SIZE_CHECK(stamp_data.frame, w, h)) {
/* extra space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
x - BUFF_MARGIN_X, y - BUFF_MARGIN_Y, x + w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
/* and pad the text. */
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.frame);
/* space width. */
x += w + pad;
}
if (TEXT_SIZE_CHECK(stamp_data.camera, w, h)) {
/* extra space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
x - BUFF_MARGIN_X, y - BUFF_MARGIN_Y, x + w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.camera);
/* space width. */
x += w + pad;
}
if (TEXT_SIZE_CHECK(stamp_data.cameralens, w, h)) {
/* extra space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
x - BUFF_MARGIN_X, y - BUFF_MARGIN_Y, x + w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
BLF_position(mono, x, y + y_ofs, 0.0);
BLF_draw_buffer(mono, stamp_data.cameralens);
}
2012-06-23 23:22:19 +00:00
if (TEXT_SIZE_CHECK(stamp_data.scene, w, h)) {
/* Bottom right corner, with an extra space because blenfont is too strict! */
x = width - w - 2;
/* extra space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
x - BUFF_MARGIN_X, y - BUFF_MARGIN_Y, x + w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
/* and pad the text. */
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.scene);
}
2012-06-23 23:22:19 +00:00
if (TEXT_SIZE_CHECK(stamp_data.strip, w, h)) {
/* Top right corner, with an extra space because blenfont is too strict! */
x = width - w - pad;
y = height - h;
/* extra space for background. */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
buf_rectfill_area(rect, rectf, width, height, scene->r.bg_stamp, display,
2012-11-09 09:33:28 +00:00
x - BUFF_MARGIN_X, y - BUFF_MARGIN_Y, x + w + BUFF_MARGIN_X, y + h + BUFF_MARGIN_Y);
BLF_position(mono, x, y + y_ofs, 0.0);
BugFix: [#20854] PROPERTIES STAMP: Rendering stamp flickers in output renders Blenfont was not thread safe, that is why one thread can change the font properties (size, dpi, color, etc) at the same time that the stamp draw on the image, and then the problem. To make blenfont thread safe I have to change two important things: 1) Every BLF_* function take one argument, the font id. 2) We have two new function to make font "thread safe": BLF_load_unique BLF_load_mem_unique This two function are for case like stamp, that need and own font that don't share the glyph cache, so can draw without problem in a different thread. Why the BLF_*_unique function ? Because blenfont keep only one copy of a font and keep a list of "glyph cache". Every glyph cache have size and dpi, so if two different thread access the same font at the same time, they can change value and finish with something like the stamp problem. Why don't remove the glyph cache ? Because if we do that, we finish with a font object for every size and dpi, and the stamp is really a special case that happen in the rendering process, so I really thing is better keep the glyph cache and make this two new function to handle this special case. (When I say "font object" I mean have the same freetype font multiple times just to have differents size and dpi) As Matt point we still can have one case that two thread access the BLF_*_unique function at the same time, but I am looking to fix this with some class of thread lock. For now I test and work fine, so if some one found problem, please let me know. Campbell I have to change the python api (python/generic/blf_api.c) to the new syntax, so maybe you can take a look at this.
2010-04-22 10:56:45 +00:00
BLF_draw_buffer(mono, stamp_data.strip);
}
/* cleanup the buffer. */
BLF_buffer(mono, NULL, NULL, 0, 0, 0, NULL);
#undef TEXT_SIZE_CHECK
#undef BUFF_MARGIN_X
#undef BUFF_MARGIN_Y
}
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
void BKE_imbuf_stamp_info(Scene *scene, Object *camera, struct ImBuf *ibuf)
{
struct StampData stamp_data;
if (!ibuf) return;
2012-06-23 23:22:19 +00:00
/* fill all the data values, no prefix */
stampdata(scene, camera, &stamp_data, 0);
2012-06-23 23:22:19 +00:00
if (stamp_data.file[0]) IMB_metadata_change_field(ibuf, "File", stamp_data.file);
if (stamp_data.note[0]) IMB_metadata_change_field(ibuf, "Note", stamp_data.note);
if (stamp_data.date[0]) IMB_metadata_change_field(ibuf, "Date", stamp_data.date);
if (stamp_data.marker[0]) IMB_metadata_change_field(ibuf, "Marker", stamp_data.marker);
if (stamp_data.time[0]) IMB_metadata_change_field(ibuf, "Time", stamp_data.time);
if (stamp_data.frame[0]) IMB_metadata_change_field(ibuf, "Frame", stamp_data.frame);
if (stamp_data.camera[0]) IMB_metadata_change_field(ibuf, "Camera", stamp_data.camera);
if (stamp_data.cameralens[0]) IMB_metadata_change_field(ibuf, "Lens", stamp_data.cameralens);
if (stamp_data.scene[0]) IMB_metadata_change_field(ibuf, "Scene", stamp_data.scene);
if (stamp_data.strip[0]) IMB_metadata_change_field(ibuf, "Strip", stamp_data.strip);
if (stamp_data.rendertime[0]) IMB_metadata_change_field(ibuf, "RenderTime", stamp_data.rendertime);
}
bool BKE_imbuf_alpha_test(ImBuf *ibuf)
{
int tot;
if (ibuf->rect_float) {
const float *buf = ibuf->rect_float;
for (tot = ibuf->x * ibuf->y; tot--; buf += 4) {
if (buf[3] < 1.0f) {
return true;
}
}
}
else if (ibuf->rect) {
unsigned char *buf = (unsigned char *)ibuf->rect;
for (tot = ibuf->x * ibuf->y; tot--; buf += 4) {
if (buf[3] != 255) {
return true;
}
}
}
return false;
}
/* note: imf->planes is ignored here, its assumed the image channels
* are already set */
int BKE_imbuf_write(ImBuf *ibuf, const char *name, ImageFormatData *imf)
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
{
char imtype = imf->imtype;
char compress = imf->compress;
char quality = imf->quality;
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
int ok;
if (imtype == R_IMF_IMTYPE_IRIS) {
ibuf->ftype = IMAGIC;
}
#ifdef WITH_HDR
else if (imtype == R_IMF_IMTYPE_RADHDR) {
ibuf->ftype = RADHDR;
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
}
#endif
else if (ELEM(imtype, R_IMF_IMTYPE_PNG, R_IMF_IMTYPE_FFMPEG, R_IMF_IMTYPE_H264, R_IMF_IMTYPE_THEORA, R_IMF_IMTYPE_XVID)) {
ibuf->ftype = PNG;
if (imtype == R_IMF_IMTYPE_PNG) {
if (imf->depth == R_IMF_CHAN_DEPTH_16)
ibuf->ftype |= PNG_16BIT;
ibuf->ftype |= compress;
}
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
}
#ifdef WITH_DDS
else if (imtype == R_IMF_IMTYPE_DDS) {
ibuf->ftype = DDS;
}
#endif
else if (imtype == R_IMF_IMTYPE_BMP) {
ibuf->ftype = BMP;
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
}
#ifdef WITH_TIFF
else if (imtype == R_IMF_IMTYPE_TIFF) {
ibuf->ftype = TIF;
if (imf->depth == R_IMF_CHAN_DEPTH_16)
ibuf->ftype |= TIF_16BIT;
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
}
#endif
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
#ifdef WITH_OPENEXR
else if (imtype == R_IMF_IMTYPE_OPENEXR || imtype == R_IMF_IMTYPE_MULTILAYER) {
ibuf->ftype = OPENEXR;
if (imf->depth == R_IMF_CHAN_DEPTH_16)
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
ibuf->ftype |= OPENEXR_HALF;
ibuf->ftype |= (imf->exr_codec & OPENEXR_COMPRESS);
2012-06-23 23:22:19 +00:00
if (!(imf->flag & R_IMF_FLAG_ZBUF))
ibuf->zbuf_float = NULL; /* signal for exr saving */
2012-06-23 23:22:19 +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
}
#endif
#ifdef WITH_CINEON
else if (imtype == R_IMF_IMTYPE_CINEON) {
ibuf->ftype = CINEON;
if (imf->cineon_flag & R_IMF_CINEON_FLAG_LOG) {
ibuf->ftype |= CINEON_LOG;
}
if (imf->depth == R_IMF_CHAN_DEPTH_16) {
ibuf->ftype |= CINEON_16BIT;
}
else if (imf->depth == R_IMF_CHAN_DEPTH_12) {
ibuf->ftype |= CINEON_12BIT;
}
else if (imf->depth == R_IMF_CHAN_DEPTH_10) {
ibuf->ftype |= CINEON_10BIT;
}
}
else if (imtype == R_IMF_IMTYPE_DPX) {
ibuf->ftype = DPX;
if (imf->cineon_flag & R_IMF_CINEON_FLAG_LOG) {
ibuf->ftype |= CINEON_LOG;
}
if (imf->depth == R_IMF_CHAN_DEPTH_16) {
ibuf->ftype |= CINEON_16BIT;
}
else if (imf->depth == R_IMF_CHAN_DEPTH_12) {
ibuf->ftype |= CINEON_12BIT;
}
else if (imf->depth == R_IMF_CHAN_DEPTH_10) {
ibuf->ftype |= CINEON_10BIT;
}
}
#endif
else if (imtype == R_IMF_IMTYPE_TARGA) {
ibuf->ftype = TGA;
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
}
else if (imtype == R_IMF_IMTYPE_RAWTGA) {
ibuf->ftype = RAWTGA;
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
}
#ifdef WITH_OPENJPEG
else if (imtype == R_IMF_IMTYPE_JP2) {
if (quality < 10) quality = 90;
ibuf->ftype = JP2 | quality;
2012-06-23 23:22:19 +00:00
if (imf->depth == R_IMF_CHAN_DEPTH_16) {
ibuf->ftype |= JP2_16BIT;
}
else if (imf->depth == R_IMF_CHAN_DEPTH_12) {
ibuf->ftype |= JP2_12BIT;
}
2012-06-23 23:22:19 +00:00
if (imf->jp2_flag & R_IMF_JP2_FLAG_YCC) {
ibuf->ftype |= JP2_YCC;
}
if (imf->jp2_flag & R_IMF_JP2_FLAG_CINE_PRESET) {
ibuf->ftype |= JP2_CINE;
if (imf->jp2_flag & R_IMF_JP2_FLAG_CINE_48)
ibuf->ftype |= JP2_CINE_48FPS;
}
if (imf->jp2_codec == R_IMF_JP2_CODEC_JP2)
ibuf->ftype |= JP2_JP2;
else if (imf->jp2_codec == R_IMF_JP2_CODEC_J2K)
ibuf->ftype |= JP2_J2K;
else
BLI_assert(!"Unsupported jp2 codec was specified in im_format->jp2_codec");
}
#endif
else {
/* R_IMF_IMTYPE_JPEG90, etc. default we save jpegs */
if (quality < 10) quality = 90;
ibuf->ftype = JPG | quality;
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
}
2012-06-23 23:22:19 +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
BLI_make_existing_file(name);
2012-06-23 23:22:19 +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
ok = IMB_saveiff(ibuf, name, IB_rect | IB_zbuf | IB_zbuffloat);
if (ok == 0) {
perror(name);
}
2012-06-23 23:22:19 +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
return(ok);
}
2014-06-15 12:15:28 +10:00
/* same as BKE_imbuf_write() but crappy workaround not to permanently modify
* _some_, values in the imbuf */
int BKE_imbuf_write_as(ImBuf *ibuf, const char *name, ImageFormatData *imf,
const bool save_copy)
{
ImBuf ibuf_back = *ibuf;
int ok;
/* all data is rgba anyway,
* this just controls how to save for some formats */
ibuf->planes = imf->planes;
ok = BKE_imbuf_write(ibuf, name, imf);
if (save_copy) {
/* note that we are not restoring _all_ settings */
ibuf->planes = ibuf_back.planes;
ibuf->ftype = ibuf_back.ftype;
}
return ok;
}
int BKE_imbuf_write_stamp(Scene *scene, struct Object *camera, ImBuf *ibuf, const char *name, struct ImageFormatData *imf)
{
if (scene && scene->r.stamp & R_STAMP_ALL)
BKE_imbuf_stamp_info(scene, camera, ibuf);
return BKE_imbuf_write(ibuf, name, imf);
}
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
static void image_path_makepicstring(
char *string, const char *base, const char *relbase, int frame, const char imtype,
const ImageFormatData *im_format, const short use_ext, const short use_frames)
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
{
if (string == NULL) return;
BLI_strncpy(string, base, FILE_MAX - 10); /* weak assumption */
BLI_path_abs(string, relbase);
if (use_frames)
BLI_path_frame(string, frame, 4);
Giant commit! A full detailed description of this will be done later... is several days of work. Here's a summary: Render: - Full cleanup of render code, removing *all* globals and bad level calls all over blender. Render module is now not called abusive anymore - API-fied calls to rendering - Full recode of internal render pipeline. Is now rendering tiles by default, prepared for much smarter 'bucket' render later. - Each thread now can render a full part - Renders were tested with 4 threads, goes fine, apart from some lookup tables in softshadow and AO still - Rendering is prepared to do multiple layers and passes - No single 32 bits trick in render code anymore, all 100% floats now. Writing images/movies - moved writing images to blender kernel (bye bye 'schrijfplaatje'!) - made a new Movie handle system, also in kernel. This will enable much easier use of movies in Blender PreviewRender: - Using new render API, previewrender (in buttons) now uses regular render code to generate images. - new datafile 'preview.blend.c' has the preview scenes in it - previews get rendered in exact displayed size (1 pixel = 1 pixel) 3D Preview render - new; press Pkey in 3d window, for a panel that continuously renders (pkey is for games, i know... but we dont do that in orange now!) - this render works nearly identical to buttons-preview render, so it stops rendering on any event (mouse, keyboard, etc) - on moving/scaling the panel, the render code doesn't recreate all geometry - same for shifting/panning view - all other operations (now) regenerate the full render database still. - this is WIP... but big fun, especially for simple scenes! Compositor - Using same node system as now in use for shaders, you can composit images - works pretty straightforward... needs much more options/tools and integration with rendering still - is not threaded yet, nor is so smart to only recalculate changes... will be done soon! - the "Render Result" node will get all layers/passes as output sockets - The "Output" node renders to a builtin image, which you can view in the Image window. (yes, output nodes to render-result, and to files, is on the list!) The Bad News - "Unified Render" is removed. It might come back in some stage, but this system should be built from scratch. I can't really understand this code... I expect it is not much needed, especially with advanced layer/passes control - Panorama render, Field render, Motion blur, is not coded yet... (I had to recode every single feature in render, so...!) - Lens Flare is also not back... needs total revision, might become composit effect though (using zbuffer for visibility) - Part render is gone! (well, thats obvious, its default now). - The render window is only restored with limited functionality... I am going to check first the option to render to a Image window, so Blender can become a true single-window application. :) For example, the 'Spare render buffer' (jkey) doesnt work. - Render with border, now default creates a smaller image - No zbuffers are written yet... on the todo! - Scons files and MSVC will need work to get compiling again OK... thats what I can quickly recall. Now go compiling!
2006-01-23 22:05:47 +00:00
if (use_ext)
image_path_ensure_ext(string, imtype, im_format);
}
2012-06-23 23:22:19 +00:00
void BKE_image_path_from_imformat(
char *string, const char *base, const char *relbase, int frame,
const ImageFormatData *im_format, const bool use_ext, const bool use_frames)
{
image_path_makepicstring(string, base, relbase, frame, im_format->imtype, im_format, use_ext, use_frames);
}
void BKE_image_path_from_imtype(
char *string, const char *base, const char *relbase, int frame,
const char imtype, const bool use_ext, const bool use_frames)
{
image_path_makepicstring(string, base, relbase, frame, imtype, NULL, use_ext, use_frames);
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
}
struct anim *openanim_noload(const char *name, int flags, int streamindex, char colorspace[IMA_MAX_SPACE])
{
struct anim *anim;
anim = IMB_open_anim(name, flags, streamindex, colorspace);
return anim;
}
/* used by sequencer too */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
struct anim *openanim(const char *name, int flags, int streamindex, char colorspace[IMA_MAX_SPACE])
2012-05-17 23:21:11 +00:00
{
struct anim *anim;
struct ImBuf *ibuf;
2012-06-23 23:22:19 +00:00
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
anim = IMB_open_anim(name, flags, streamindex, colorspace);
if (anim == NULL) return NULL;
2002-10-12 11:37:38 +00:00
ibuf = IMB_anim_absolute(anim, 0, IMB_TC_NONE, IMB_PROXY_NONE);
if (ibuf == NULL) {
if (BLI_exists(name))
printf("not an anim: %s\n", name);
else
printf("anim file doesn't exist: %s\n", name);
2002-10-12 11:37:38 +00:00
IMB_free_anim(anim);
return NULL;
2002-10-12 11:37:38 +00:00
}
IMB_freeImBuf(ibuf);
2012-06-23 23:22:19 +00:00
2002-10-12 11:37:38 +00:00
return(anim);
}
/* ************************* New Image API *************** */
2012-06-23 23:22:19 +00:00
/* Notes about Image storage
* - packedfile
* -> written in .blend
* - filename
* -> written in .blend
* - movie
* -> comes from packedfile or filename
* - renderresult
* -> comes from packedfile or filename
* - listbase
* -> ibufs from exrhandle
* - flipbook array
* -> ibufs come from movie, temporary renderresult or sequence
* - ibuf
* -> comes from packedfile or filename or generated
*/
/* forces existence of 1 Image for renderout or nodes, returns Image */
/* name is only for default, when making new one */
Image *BKE_image_verify_viewer(int type, const char *name)
{
Image *ima;
2012-06-23 23:22:19 +00:00
for (ima = G.main->image.first; ima; ima = ima->id.next)
if (ima->source == IMA_SRC_VIEWER)
if (ima->type == type)
break;
2012-06-23 23:22:19 +00:00
if (ima == NULL)
ima = image_alloc(G.main, name, IMA_SRC_VIEWER, type);
2012-06-23 23:22:19 +00:00
/* happens on reload, imagewindow cannot be image user when hidden*/
if (ima->id.us == 0)
id_us_plus(&ima->id);
2002-10-12 11:37:38 +00:00
return ima;
}
2002-10-12 11:37:38 +00:00
void BKE_image_walk_all_users(const Main *mainp, void *customdata,
void callback(Image *ima, ImageUser *iuser, void *customdata))
{
wmWindowManager *wm;
wmWindow *win;
Tex *tex;
/* texture users */
for (tex = mainp->tex.first; tex; tex = tex->id.next) {
if (tex->type == TEX_IMAGE && tex->ima) {
callback(tex->ima, &tex->iuser, customdata);
}
}
/* image window, compo node users */
for (wm = mainp->wm.first; wm; wm = wm->id.next) { /* only 1 wm */
for (win = wm->windows.first; win; win = win->next) {
ScrArea *sa;
for (sa = win->screen->areabase.first; sa; sa = sa->next) {
if (sa->spacetype == SPACE_VIEW3D) {
View3D *v3d = sa->spacedata.first;
BGpic *bgpic;
for (bgpic = v3d->bgpicbase.first; bgpic; bgpic = bgpic->next) {
callback(bgpic->ima, &bgpic->iuser, customdata);
}
}
else if (sa->spacetype == SPACE_IMAGE) {
SpaceImage *sima = sa->spacedata.first;
callback(sima->image, &sima->iuser, customdata);
}
else if (sa->spacetype == SPACE_NODE) {
SpaceNode *snode = sa->spacedata.first;
2013-03-18 18:25:05 +00:00
if (snode->nodetree && snode->nodetree->type == NTREE_COMPOSIT) {
bNode *node;
for (node = snode->nodetree->nodes.first; node; node = node->next) {
if (node->id && node->type == CMP_NODE_IMAGE) {
Image *ima = (Image *)node->id;
ImageUser *iuser = node->storage;
callback(ima, iuser, customdata);
}
}
}
}
}
}
}
}
static void image_tag_frame_recalc(Image *ima, ImageUser *iuser, void *customdata)
{
Image *changed_image = customdata;
if (ima == changed_image && BKE_image_is_animated(ima)) {
iuser->flag |= IMA_NEED_FRAME_RECALC;
}
}
void BKE_image_signal(Image *ima, ImageUser *iuser, int signal)
{
if (ima == NULL)
return;
2012-06-23 23:22:19 +00:00
BLI_spin_lock(&image_spin);
switch (signal) {
case IMA_SIGNAL_FREE:
BKE_image_free_buffers(ima);
if (iuser)
iuser->ok = 1;
break;
case IMA_SIGNAL_SRC_CHANGE:
if (ima->type == IMA_TYPE_UV_TEST)
if (ima->source != IMA_SRC_GENERATED)
ima->type = IMA_TYPE_IMAGE;
if (ima->source == IMA_SRC_GENERATED) {
if (ima->gen_x == 0 || ima->gen_y == 0) {
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ImBuf *ibuf = image_get_cached_ibuf_for_index_frame(ima, IMA_NO_INDEX, 0);
if (ibuf) {
ima->gen_x = ibuf->x;
ima->gen_y = ibuf->y;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
IMB_freeImBuf(ibuf);
}
}
/* Changing source type to generated will likely change file format
* used by generated image buffer. Saving different file format to
* the old name might confuse other applications.
*
* Here we ensure original image path wouldn't be used when saving
* generated image.
*/
ima->name[0] = '\0';
}
#if 0
/* force reload on first use, but not for multilayer, that makes nodes and buttons in ui drawing fail */
if (ima->type != IMA_TYPE_MULTILAYER)
2014-05-22 17:40:35 +10:00
BKE_image_free_buffers(ima);
#else
/* image buffers for non-sequence multilayer will share buffers with RenderResult,
* however sequence multilayer will own buffers. Such logic makes switching from
2013-10-31 14:10:01 +00:00
* single multilayer file to sequence completely unstable
* since changes in nodes seems this workaround isn't needed anymore, all sockets
* are nicely detecting anyway, but freeing buffers always here makes multilayer
* sequences behave stable
*/
2014-05-22 17:40:35 +10:00
BKE_image_free_buffers(ima);
#endif
ima->ok = 1;
if (iuser)
iuser->ok = 1;
BKE_image_walk_all_users(G.main, ima, image_tag_frame_recalc);
break;
case IMA_SIGNAL_RELOAD:
/* try to repack file */
if (ima->packedfile) {
PackedFile *pf;
pf = newPackedFile(NULL, ima->name, ID_BLEND_PATH(G.main, &ima->id));
if (pf) {
freePackedFile(ima->packedfile);
ima->packedfile = pf;
2014-05-22 17:40:35 +10:00
BKE_image_free_buffers(ima);
}
else {
printf("ERROR: Image not available. Keeping packed image\n");
}
}
else
2014-05-22 17:40:35 +10:00
BKE_image_free_buffers(ima);
2012-06-23 23:22:19 +00:00
if (iuser)
iuser->ok = 1;
2012-06-23 23:22:19 +00:00
break;
case IMA_SIGNAL_USER_NEW_IMAGE:
if (iuser) {
iuser->ok = 1;
if (ima->source == IMA_SRC_FILE || ima->source == IMA_SRC_SEQUENCE) {
if (ima->type == IMA_TYPE_MULTILAYER) {
iuser->multi_index = 0;
iuser->layer = iuser->pass = 0;
}
}
}
break;
case IMA_SIGNAL_COLORMANAGE:
BKE_image_free_buffers(ima);
ima->ok = 1;
if (iuser)
iuser->ok = 1;
break;
2002-10-12 11:37:38 +00:00
}
2012-06-23 23:22:19 +00:00
BLI_spin_unlock(&image_spin);
2012-03-18 07:38:51 +00:00
/* don't use notifiers because they are not 100% sure to succeeded
* this also makes sure all scenes are accounted for. */
{
Scene *scene;
for (scene = G.main->scene.first; scene; scene = scene->id.next) {
if (scene->nodetree) {
nodeUpdateID(scene->nodetree, &ima->id);
}
}
}
2002-10-12 11:37:38 +00:00
}
/* if layer or pass changes, we need an index for the imbufs list */
/* note it is called for rendered results, but it doesnt use the index! */
/* and because rendered results use fake layer/passes, don't correct for wrong indices here */
RenderPass *BKE_image_multilayer_index(RenderResult *rr, ImageUser *iuser)
2002-10-12 11:37:38 +00:00
{
RenderLayer *rl;
RenderPass *rpass = NULL;
2012-06-23 23:22:19 +00:00
if (rr == NULL)
return NULL;
2012-06-23 23:22:19 +00:00
if (iuser) {
short index = 0, rl_index = 0, rp_index;
2012-06-23 23:22:19 +00:00
for (rl = rr->layers.first; rl; rl = rl->next, rl_index++) {
rp_index = 0;
for (rpass = rl->passes.first; rpass; rpass = rpass->next, index++, rp_index++)
if (iuser->layer == rl_index && iuser->pass == rp_index)
break;
if (rpass)
break;
}
2012-06-23 23:22:19 +00:00
if (rpass)
iuser->multi_index = index;
2012-06-23 23:22:19 +00:00
else
iuser->multi_index = 0;
}
if (rpass == NULL) {
rl = rr->layers.first;
if (rl)
rpass = rl->passes.first;
}
2012-06-23 23:22:19 +00:00
return rpass;
}
RenderResult *BKE_image_acquire_renderresult(Scene *scene, Image *ima)
{
if (ima->rr) {
return ima->rr;
}
else if (ima->type == IMA_TYPE_R_RESULT) {
if (ima->render_slot == ima->last_render_slot)
return RE_AcquireResultRead(RE_GetRender(scene->id.name));
else
return ima->renders[ima->render_slot];
}
else
return NULL;
}
void BKE_image_release_renderresult(Scene *scene, Image *ima)
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
{
if (ima->rr) {
/* pass */
}
else if (ima->type == IMA_TYPE_R_RESULT) {
if (ima->render_slot == ima->last_render_slot)
RE_ReleaseResult(RE_GetRender(scene->id.name));
}
}
void BKE_image_backup_render(Scene *scene, Image *ima)
{
/* called right before rendering, ima->renders contains render
* result pointers for everything but the current render */
Render *re = RE_GetRender(scene->id.name);
int slot = ima->render_slot, last = ima->last_render_slot;
if (slot != last) {
if (ima->renders[slot]) {
RE_FreeRenderResult(ima->renders[slot]);
ima->renders[slot] = NULL;
}
ima->renders[last] = NULL;
RE_SwapResult(re, &ima->renders[last]);
}
ima->last_render_slot = slot;
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
}
/* after imbuf load, openexr type can return with a exrhandle open */
/* in that case we have to build a render-result */
static void image_create_multilayer(Image *ima, ImBuf *ibuf, int framenr)
{
const char *colorspace = ima->colorspace_settings.name;
bool predivide = (ima->alpha_mode == IMA_ALPHA_PREMUL);
2012-06-23 23:22:19 +00:00
ima->rr = RE_MultilayerConvert(ibuf->userdata, colorspace, predivide, ibuf->x, ibuf->y);
#ifdef WITH_OPENEXR
IMB_exr_close(ibuf->userdata);
#endif
ibuf->userdata = NULL;
if (ima->rr)
ima->rr->framenr = framenr;
}
/* common stuff to do with images after loading */
static void image_initialize_after_load(Image *ima, ImBuf *ibuf)
{
/* preview is NULL when it has never been used as an icon before */
if (G.background == 0 && ima->preview == NULL)
BKE_icon_changed(BKE_icon_getid(&ima->id));
/* fields */
if (ima->flag & IMA_FIELDS) {
if (ima->flag & IMA_STD_FIELD) de_interlace_st(ibuf);
else de_interlace_ng(ibuf);
}
/* timer */
BKE_image_tag_time(ima);
2012-06-23 23:22:19 +00:00
ima->ok = IMA_OK_LOADED;
2012-06-23 23:22:19 +00:00
}
Alpha premul pipeline cleanup This assumptions are now made: - Internally float buffers are always linear alpha-premul colors - Readers should worry about delivering float buffers with that assumptions. - There's an input image setting to say whether it's stored with straight/premul alpha on the disk. - Byte buffers are now assumed have straight alpha, readers should deliver straight alpha. Some implementation details: - Removed scene's color unpremultiply setting, which was very much confusing and was wrong for default settings. Now all renderers assumes to deliver premultiplied alpha. - IMB_buffer_byte_from_float will now linearize alpha when converting from buffer. - Sequencer's effects were changed to assume bytes have got straight alpha. Most of effects will work with bytes still, however for glow it was more tricky to avoid data loss, so there's a commented out glow implementation which converts byte buffer to floats first, operates on floats and returns bytes back. It's slower and not sure if it should actually be used -- who're using glow on alpha anyway? - Sequencer modifiers should also be working nice with straight bytes now. - GLSL preview will predivide float textures to make nice shading, shading with byte textures worked nice (GLSL was assuming straight alpha). - Blender Internal will set alpha=1 to the whole sky. The same happens in Cycles and there's no way to avoid this -- sky is neither straight nor premul and doesn't fit color pipeline well. - Straight alpha mode for render result was also eliminated. - Conversion to correct alpha need to be done before linearizing float buffer. - TIFF will now load and save files with proper alpha mode setting in file meta data header. - Remove Use Alpha from texture mapping and replaced with image datablock setting. Behaves much more predictable and clear from code point of view and solves possible regressions when non-premultiplied images were used as textures with ignoring alpha channel.
2012-12-31 13:52:13 +00:00
static int imbuf_alpha_flags_for_image(Image *ima)
{
int flag = 0;
if (ima->flag & IMA_IGNORE_ALPHA)
flag |= IB_ignore_alpha;
else if (ima->alpha_mode == IMA_ALPHA_PREMUL)
flag |= IB_alphamode_premul;
return flag;
}
static ImBuf *image_load_sequence_file(Image *ima, ImageUser *iuser, int frame)
{
struct ImBuf *ibuf;
char name[FILE_MAX];
int flag;
2012-06-23 23:22:19 +00:00
/* XXX temp stuff? */
if (ima->lastframe != frame)
ima->tpageflag |= IMA_TPAGE_REFRESH;
ima->lastframe = frame;
BKE_image_user_file_path(iuser, ima, name);
2012-06-23 23:22:19 +00:00
flag = IB_rect | IB_multilayer;
Alpha premul pipeline cleanup This assumptions are now made: - Internally float buffers are always linear alpha-premul colors - Readers should worry about delivering float buffers with that assumptions. - There's an input image setting to say whether it's stored with straight/premul alpha on the disk. - Byte buffers are now assumed have straight alpha, readers should deliver straight alpha. Some implementation details: - Removed scene's color unpremultiply setting, which was very much confusing and was wrong for default settings. Now all renderers assumes to deliver premultiplied alpha. - IMB_buffer_byte_from_float will now linearize alpha when converting from buffer. - Sequencer's effects were changed to assume bytes have got straight alpha. Most of effects will work with bytes still, however for glow it was more tricky to avoid data loss, so there's a commented out glow implementation which converts byte buffer to floats first, operates on floats and returns bytes back. It's slower and not sure if it should actually be used -- who're using glow on alpha anyway? - Sequencer modifiers should also be working nice with straight bytes now. - GLSL preview will predivide float textures to make nice shading, shading with byte textures worked nice (GLSL was assuming straight alpha). - Blender Internal will set alpha=1 to the whole sky. The same happens in Cycles and there's no way to avoid this -- sky is neither straight nor premul and doesn't fit color pipeline well. - Straight alpha mode for render result was also eliminated. - Conversion to correct alpha need to be done before linearizing float buffer. - TIFF will now load and save files with proper alpha mode setting in file meta data header. - Remove Use Alpha from texture mapping and replaced with image datablock setting. Behaves much more predictable and clear from code point of view and solves possible regressions when non-premultiplied images were used as textures with ignoring alpha channel.
2012-12-31 13:52:13 +00:00
flag |= imbuf_alpha_flags_for_image(ima);
/* read ibuf */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
ibuf = IMB_loadiffname(name, flag, ima->colorspace_settings.name);
#if 0
if (ibuf) {
printf(AT " loaded %s\n", name);
}
else {
printf(AT " missed %s\n", name);
}
#endif
if (ibuf) {
#ifdef WITH_OPENEXR
/* handle multilayer case, don't assign ibuf. will be handled in BKE_image_acquire_ibuf */
if (ibuf->ftype == OPENEXR && ibuf->userdata) {
2012-06-23 23:22:19 +00:00
image_create_multilayer(ima, ibuf, frame);
ima->type = IMA_TYPE_MULTILAYER;
IMB_freeImBuf(ibuf);
ibuf = NULL;
}
else {
image_initialize_after_load(ima, ibuf);
image_assign_ibuf(ima, ibuf, 0, frame);
}
#else
image_initialize_after_load(ima, ibuf);
image_assign_ibuf(ima, ibuf, 0, frame);
#endif
}
else
ima->ok = 0;
2012-06-23 23:22:19 +00:00
if (iuser)
iuser->ok = ima->ok;
2012-06-23 23:22:19 +00:00
return ibuf;
}
static ImBuf *image_load_sequence_multilayer(Image *ima, ImageUser *iuser, int frame)
{
struct ImBuf *ibuf = NULL;
2012-06-23 23:22:19 +00:00
/* either we load from RenderResult, or we have to load a new one */
2012-06-23 23:22:19 +00:00
/* check for new RenderResult */
if (ima->rr == NULL || frame != ima->rr->framenr) {
if (ima->rr) {
/* Cached image buffers shares pointers with render result,
* need to ensure there's no image buffers are hanging around
* with dead links after freeing the render result.
*/
2014-10-17 12:37:40 +02:00
image_free_cached_frames(ima);
RE_FreeRenderResult(ima->rr);
ima->rr = NULL;
}
2012-06-23 23:22:19 +00:00
ibuf = image_load_sequence_file(ima, iuser, frame);
2012-06-23 23:22:19 +00:00
if (ibuf) { /* actually an error */
ima->type = IMA_TYPE_IMAGE;
printf("error, multi is normal image\n");
}
}
if (ima->rr) {
RenderPass *rpass = BKE_image_multilayer_index(ima->rr, iuser);
2012-06-23 23:22:19 +00:00
if (rpass) {
// printf("load from pass %s\n", rpass->name);
/* since we free render results, we copy the rect */
ibuf = IMB_allocImBuf(ima->rr->rectx, ima->rr->recty, 32, 0);
ibuf->rect_float = MEM_dupallocN(rpass->rect);
ibuf->flags |= IB_rectfloat;
ibuf->mall = IB_rectfloat;
ibuf->channels = rpass->channels;
2012-06-23 23:22:19 +00:00
image_initialize_after_load(ima, ibuf);
image_assign_ibuf(ima, ibuf, iuser ? iuser->multi_index : 0, frame);
2012-06-23 23:22:19 +00:00
}
// else printf("pass not found\n");
2002-10-12 11:37:38 +00:00
}
else
ima->ok = 0;
2012-06-23 23:22:19 +00:00
if (iuser)
iuser->ok = ima->ok;
2012-06-23 23:22:19 +00:00
return ibuf;
2002-10-12 11:37:38 +00:00
}
static ImBuf *image_load_movie_file(Image *ima, ImageUser *iuser, int frame)
2002-10-12 11:37:38 +00:00
{
struct ImBuf *ibuf = NULL;
2012-06-23 23:22:19 +00:00
ima->lastframe = frame;
2012-06-23 23:22:19 +00:00
if (ima->anim == NULL) {
char str[FILE_MAX];
int flags = IB_rect;
if (ima->flag & IMA_DEINTERLACE) {
flags |= IB_animdeinterlace;
}
2012-06-23 23:22:19 +00:00
BKE_image_user_file_path(iuser, ima, str);
/* FIXME: make several stream accessible in image editor, too*/
ima->anim = openanim(str, flags, 0, ima->colorspace_settings.name);
2012-06-23 23:22:19 +00:00
/* let's initialize this user */
if (ima->anim && iuser && iuser->frames == 0)
iuser->frames = IMB_anim_get_duration(ima->anim,
IMB_TC_RECORD_RUN);
}
2012-06-23 23:22:19 +00:00
if (ima->anim) {
int dur = IMB_anim_get_duration(ima->anim,
IMB_TC_RECORD_RUN);
int fra = frame - 1;
2012-06-23 23:22:19 +00:00
if (fra < 0) fra = 0;
if (fra > (dur - 1)) fra = dur - 1;
ibuf = IMB_makeSingleUser(
IMB_anim_absolute(ima->anim, fra,
IMB_TC_RECORD_RUN,
IMB_PROXY_NONE));
2012-06-23 23:22:19 +00:00
if (ibuf) {
image_initialize_after_load(ima, ibuf);
image_assign_ibuf(ima, ibuf, 0, frame);
}
else
ima->ok = 0;
2002-10-12 11:37:38 +00:00
}
else
ima->ok = 0;
2012-06-23 23:22:19 +00:00
if (iuser)
iuser->ok = ima->ok;
2012-06-23 23:22:19 +00:00
return ibuf;
2002-10-12 11:37:38 +00:00
}
/* warning, 'iuser' can be NULL */
static ImBuf *image_load_image_file(Image *ima, ImageUser *iuser, int cfra)
{
struct ImBuf *ibuf;
char str[FILE_MAX];
int assign = 0, flag;
2012-06-23 23:22:19 +00:00
/* always ensure clean ima */
2014-05-22 17:40:35 +10:00
BKE_image_free_buffers(ima);
2012-06-23 23:22:19 +00:00
/* is there a PackedFile with this image ? */
if (ima->packedfile) {
flag = IB_rect | IB_multilayer;
Alpha premul pipeline cleanup This assumptions are now made: - Internally float buffers are always linear alpha-premul colors - Readers should worry about delivering float buffers with that assumptions. - There's an input image setting to say whether it's stored with straight/premul alpha on the disk. - Byte buffers are now assumed have straight alpha, readers should deliver straight alpha. Some implementation details: - Removed scene's color unpremultiply setting, which was very much confusing and was wrong for default settings. Now all renderers assumes to deliver premultiplied alpha. - IMB_buffer_byte_from_float will now linearize alpha when converting from buffer. - Sequencer's effects were changed to assume bytes have got straight alpha. Most of effects will work with bytes still, however for glow it was more tricky to avoid data loss, so there's a commented out glow implementation which converts byte buffer to floats first, operates on floats and returns bytes back. It's slower and not sure if it should actually be used -- who're using glow on alpha anyway? - Sequencer modifiers should also be working nice with straight bytes now. - GLSL preview will predivide float textures to make nice shading, shading with byte textures worked nice (GLSL was assuming straight alpha). - Blender Internal will set alpha=1 to the whole sky. The same happens in Cycles and there's no way to avoid this -- sky is neither straight nor premul and doesn't fit color pipeline well. - Straight alpha mode for render result was also eliminated. - Conversion to correct alpha need to be done before linearizing float buffer. - TIFF will now load and save files with proper alpha mode setting in file meta data header. - Remove Use Alpha from texture mapping and replaced with image datablock setting. Behaves much more predictable and clear from code point of view and solves possible regressions when non-premultiplied images were used as textures with ignoring alpha channel.
2012-12-31 13:52:13 +00:00
flag |= imbuf_alpha_flags_for_image(ima);
2012-06-23 23:22:19 +00:00
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
ibuf = IMB_ibImageFromMemory((unsigned char *)ima->packedfile->data, ima->packedfile->size, flag,
ima->colorspace_settings.name, "<packed data>");
2012-06-23 23:22:19 +00:00
}
else {
flag = IB_rect | IB_multilayer | IB_metadata;
Alpha premul pipeline cleanup This assumptions are now made: - Internally float buffers are always linear alpha-premul colors - Readers should worry about delivering float buffers with that assumptions. - There's an input image setting to say whether it's stored with straight/premul alpha on the disk. - Byte buffers are now assumed have straight alpha, readers should deliver straight alpha. Some implementation details: - Removed scene's color unpremultiply setting, which was very much confusing and was wrong for default settings. Now all renderers assumes to deliver premultiplied alpha. - IMB_buffer_byte_from_float will now linearize alpha when converting from buffer. - Sequencer's effects were changed to assume bytes have got straight alpha. Most of effects will work with bytes still, however for glow it was more tricky to avoid data loss, so there's a commented out glow implementation which converts byte buffer to floats first, operates on floats and returns bytes back. It's slower and not sure if it should actually be used -- who're using glow on alpha anyway? - Sequencer modifiers should also be working nice with straight bytes now. - GLSL preview will predivide float textures to make nice shading, shading with byte textures worked nice (GLSL was assuming straight alpha). - Blender Internal will set alpha=1 to the whole sky. The same happens in Cycles and there's no way to avoid this -- sky is neither straight nor premul and doesn't fit color pipeline well. - Straight alpha mode for render result was also eliminated. - Conversion to correct alpha need to be done before linearizing float buffer. - TIFF will now load and save files with proper alpha mode setting in file meta data header. - Remove Use Alpha from texture mapping and replaced with image datablock setting. Behaves much more predictable and clear from code point of view and solves possible regressions when non-premultiplied images were used as textures with ignoring alpha channel.
2012-12-31 13:52:13 +00:00
flag |= imbuf_alpha_flags_for_image(ima);
2012-06-23 23:22:19 +00:00
/* get the right string */
BKE_image_user_frame_calc(iuser, cfra, 0);
BKE_image_user_file_path(iuser, ima, str);
2012-06-23 23:22:19 +00:00
/* read ibuf */
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
ibuf = IMB_loadiffname(str, flag, ima->colorspace_settings.name);
}
2012-06-23 23:22:19 +00:00
if (ibuf) {
/* handle multilayer case, don't assign ibuf. will be handled in BKE_image_acquire_ibuf */
if (ibuf->ftype == OPENEXR && ibuf->userdata) {
2012-06-23 23:22:19 +00:00
image_create_multilayer(ima, ibuf, cfra);
ima->type = IMA_TYPE_MULTILAYER;
IMB_freeImBuf(ibuf);
ibuf = NULL;
}
else {
image_initialize_after_load(ima, ibuf);
assign = 1;
/* check if the image is a font image... */
detectBitmapFont(ibuf);
2012-06-23 23:22:19 +00:00
/* make packed file for autopack */
if ((ima->packedfile == NULL) && (G.fileflags & G_AUTOPACK))
ima->packedfile = newPackedFile(NULL, str, ID_BLEND_PATH(G.main, &ima->id));
}
}
else
ima->ok = 0;
2012-06-23 23:22:19 +00:00
if (assign)
image_assign_ibuf(ima, ibuf, IMA_NO_INDEX, 0);
if (iuser)
iuser->ok = ima->ok;
2012-06-23 23:22:19 +00:00
return ibuf;
}
static ImBuf *image_get_ibuf_multilayer(Image *ima, ImageUser *iuser)
2002-10-12 11:37:38 +00:00
{
ImBuf *ibuf = NULL;
2012-06-23 23:22:19 +00:00
if (ima->rr == NULL) {
ibuf = image_load_image_file(ima, iuser, 0);
if (ibuf) { /* actually an error */
ima->type = IMA_TYPE_IMAGE;
return ibuf;
}
}
if (ima->rr) {
RenderPass *rpass = BKE_image_multilayer_index(ima->rr, iuser);
if (rpass) {
ibuf = IMB_allocImBuf(ima->rr->rectx, ima->rr->recty, 32, 0);
2012-06-23 23:22:19 +00:00
image_initialize_after_load(ima, ibuf);
2012-06-23 23:22:19 +00:00
ibuf->rect_float = rpass->rect;
ibuf->flags |= IB_rectfloat;
ibuf->channels = rpass->channels;
image_assign_ibuf(ima, ibuf, iuser ? iuser->multi_index : IMA_NO_INDEX, 0);
}
}
2012-06-23 23:22:19 +00:00
if (ibuf == NULL)
ima->ok = 0;
if (iuser)
iuser->ok = ima->ok;
2012-06-23 23:22:19 +00:00
return ibuf;
}
/* showing RGBA result itself (from compo/sequence) or
* like exr, using layers etc */
/* always returns a single ibuf, also during render progress */
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
static ImBuf *image_get_render_result(Image *ima, ImageUser *iuser, void **lock_r)
{
Render *re;
RenderResult rres;
float *rectf, *rectz;
unsigned int *rect;
float dither;
int channels, layer, pass;
ImBuf *ibuf;
int from_render = (ima->render_slot == ima->last_render_slot);
bool byte_buffer_in_display_space = false;
if (!(iuser && iuser->scene))
return NULL;
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
/* if we the caller is not going to release the lock, don't give the image */
if (!lock_r)
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
return NULL;
re = RE_GetRender(iuser->scene->id.name);
channels = 4;
layer = iuser->layer;
pass = iuser->pass;
if (from_render) {
RE_AcquireResultImage(re, &rres);
}
else if (ima->renders[ima->render_slot]) {
rres = *(ima->renders[ima->render_slot]);
rres.have_combined = rres.rectf != NULL;
}
else
memset(&rres, 0, sizeof(RenderResult));
2012-06-23 23:22:19 +00:00
if (!(rres.rectx > 0 && rres.recty > 0)) {
if (from_render)
RE_ReleaseResultImage(re);
return NULL;
}
/* release is done in BKE_image_release_ibuf using lock_r */
if (from_render) {
BLI_lock_thread(LOCK_VIEWER);
*lock_r = re;
}
/* this gives active layer, composite or sequence result */
rect = (unsigned int *)rres.rect32;
rectf = rres.rectf;
rectz = rres.rectz;
dither = iuser->scene->r.dither_intensity;
/* combined layer gets added as first layer */
if (rres.have_combined && layer == 0) {
/* pass */
}
else if (rect && layer == 0) {
/* rect32 is set when there's a Sequence pass, this pass seems
* to have layer=0 (this is from image_buttons.c)
* in this case we ignore float buffer, because it could have
* hung from previous pass which was float
*/
rectf = NULL;
}
else if (rres.layers.first) {
RenderLayer *rl = BLI_findlink(&rres.layers, layer - (rres.have_combined ? 1 : 0));
if (rl) {
RenderPass *rpass;
/* there's no combined pass, is in renderlayer itself */
if (pass == 0) {
rectf = rl->rectf;
if (rectf == NULL) {
/* Happens when Save Buffers is enabled.
* Use display buffer stored in the render layer.
*/
rect = (unsigned int *) rl->display_buffer;
byte_buffer_in_display_space = true;
}
}
else {
rpass = BLI_findlink(&rl->passes, pass - 1);
if (rpass) {
channels = rpass->channels;
rectf = rpass->rect;
dither = 0.0f; /* don't dither passes */
}
}
for (rpass = rl->passes.first; rpass; rpass = rpass->next)
if (rpass->passtype == SCE_PASS_Z)
rectz = rpass->rect;
}
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf_for_index_frame(ima, IMA_NO_INDEX, 0);
/* make ibuf if needed, and initialize it */
if (ibuf == NULL) {
ibuf = IMB_allocImBuf(rres.rectx, rres.recty, 32, 0);
image_assign_ibuf(ima, ibuf, IMA_NO_INDEX, 0);
}
/* Set color space settings for a byte buffer.
*
* This is mainly to make it so color management treats byte buffer
* from render result with Save Buffers enabled as final display buffer
* and doesnt' apply any color management on it.
*
* For other cases we need to be sure it stays to default byte buffer space.
*/
if (ibuf->rect != rect) {
if (byte_buffer_in_display_space) {
const char *colorspace =
IMB_colormanagement_get_display_colorspace_name(&iuser->scene->view_settings,
&iuser->scene->display_settings);
IMB_colormanagement_assign_rect_colorspace(ibuf, colorspace);
}
else {
const char *colorspace = IMB_colormanagement_role_colorspace_name_get(COLOR_ROLE_DEFAULT_BYTE);
IMB_colormanagement_assign_rect_colorspace(ibuf, colorspace);
}
}
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
/* invalidate color managed buffers if render result changed */
BLI_lock_thread(LOCK_COLORMANAGE);
if (ibuf->x != rres.rectx || ibuf->y != rres.recty || ibuf->rect_float != rectf) {
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
ibuf->userflags |= IB_DISPLAY_BUFFER_INVALID;
}
ibuf->x = rres.rectx;
ibuf->y = rres.recty;
2012-06-23 23:22:19 +00:00
if (rect) {
Changes to partial update during rendering Summary: Mainly addressed to solve old TODO with color managed fallback to CPU mode when displaying render result during rendering. That fallback was caused by the fact that partial image update was always acquiring image buffer for composite output and was only modifying display buffer directly. This was a big issue for Cycles rendering which renders layers one by one and wanted to display progress of each individual layer. This lead to situations when display buffer was based on what Cycles passes via RenderResult and didn't take layer/pass from image editor header into account. Now made it so image buffer which partial update is operating with always corresponds to what is set in image editor header. To make Cycles displaying progress of all the layers one by one made it so image_rect_update switches image editor user to newly rendering render layer. It happens only once when render engine starts rendering next render layer, so should not be annoying for navigation during rendering. Additional change to render engines was done to make it so they're able to merge composite output to final result without marking tile as done. This is done via do_merge_result argument to end_result() callback. This argument is optional so should not break script compatibility. Additional changes: - Partial display update for Blender Internal now happens from the same thread as tile rendering. This makes it so display conversion (which could be pretty heavy actually) is done in separate threads. Also gives better UI feedback when rendering easy scene with small tiles. - Avoid freeing/allocating byte buffer for render result if it's owned by the image buffer. Only mark it as invalid for color management. Saves loads of buffer re-allocations in cases when having several image editors opened with render result. This change in conjunction with the rest of the patch gave around 50%-100% speedup of render time when displaying non-combined pass during rendering on my laptop. - Partial display buffer update was wrong for buffers with number of channels different from 4. - Remove unused window from RenderJob. - Made image_buffer_rect_update static since it's only used in single file. Reviewers: brecht Reviewed By: brecht CC: dingto Differential Revision: http://developer.blender.org/D98
2013-12-17 23:42:38 +06:00
imb_freerectImBuf(ibuf);
ibuf->rect = rect;
}
else {
/* byte buffer of render result has been freed, make sure image buffers
* does not reference to this buffer anymore
* need check for whether byte buffer was allocated and owned by image itself
* or if it's reusing buffer from render result
*/
if ((ibuf->mall & IB_rect) == 0)
ibuf->rect = NULL;
}
2012-06-23 23:22:19 +00:00
if (rectf) {
ibuf->rect_float = rectf;
ibuf->flags |= IB_rectfloat;
ibuf->channels = channels;
}
else {
ibuf->rect_float = NULL;
ibuf->flags &= ~IB_rectfloat;
}
if (rectz) {
ibuf->zbuf_float = rectz;
ibuf->flags |= IB_zbuffloat;
}
else {
ibuf->zbuf_float = NULL;
ibuf->flags &= ~IB_zbuffloat;
}
Color Management, Stage 2: Switch color pipeline to use OpenColorIO Replace old color pipeline which was supporting linear/sRGB color spaces only with OpenColorIO-based pipeline. This introduces two configurable color spaces: - Input color space for images and movie clips. This space is used to convert images/movies from color space in which file is saved to Blender's linear space (for float images, byte images are not internally converted, only input space is stored for such images and used later). This setting could be found in image/clip data block settings. - Display color space which defines space in which particular display is working. This settings could be found in scene's Color Management panel. When render result is being displayed on the screen, apart from converting image to display space, some additional conversions could happen. This conversions are: - View, which defines tone curve applying before display transformation. These are different ways to view the image on the same display device. For example it could be used to emulate film view on sRGB display. - Exposure affects on image exposure before tone map is applied. - Gamma is post-display gamma correction, could be used to match particular display gamma. - RGB curves are user-defined curves which are applying before display transformation, could be used for different purposes. All this settings by default are only applying on render result and does not affect on other images. If some particular image needs to be affected by this transformation, "View as Render" setting of image data block should be set to truth. Movie clips are always affected by all display transformations. This commit also introduces configurable color space in which sequencer is working. This setting could be found in scene's Color Management panel and it should be used if such stuff as grading needs to be done in color space different from sRGB (i.e. when Film view on sRGB display is use, using VD16 space as sequencer's internal space would make grading working in space which is close to the space using for display). Some technical notes: - Image buffer's float buffer is now always in linear space, even if it was created from 16bit byte images. - Space of byte buffer is stored in image buffer's rect_colorspace property. - Profile of image buffer was removed since it's not longer meaningful. - OpenGL and GLSL is supposed to always work in sRGB space. It is possible to support other spaces, but it's quite large project which isn't so much important. - Legacy Color Management option disabled is emulated by using None display. It could have some regressions, but there's no clear way to avoid them. - If OpenColorIO is disabled on build time, it should make blender behaving in the same way as previous release with color management enabled. More details could be found at this page (more details would be added soon): http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.64/Color_Management -- Thanks to Xavier Thomas, Lukas Toene for initial work on OpenColorIO integration and to Brecht van Lommel for some further development and code/ usecase review!
2012-09-15 10:05:07 +00:00
BLI_unlock_thread(LOCK_COLORMANAGE);
ibuf->dither = dither;
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
ima->ok = IMA_OK_LOADED;
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
return ibuf;
}
static void image_get_frame_and_index(Image *ima, ImageUser *iuser, int *r_frame, int *r_index)
{
int frame = 0, index = 0;
/* see if we already have an appropriate ibuf, with image source and type */
if (ima->source == IMA_SRC_MOVIE) {
frame = iuser ? iuser->framenr : ima->lastframe;
}
else if (ima->source == IMA_SRC_SEQUENCE) {
if (ima->type == IMA_TYPE_IMAGE) {
frame = iuser ? iuser->framenr : ima->lastframe;
}
else if (ima->type == IMA_TYPE_MULTILAYER) {
frame = iuser ? iuser->framenr : ima->lastframe;
index = iuser ? iuser->multi_index : IMA_NO_INDEX;
}
}
*r_frame = frame;
*r_index = index;
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
/* Get the ibuf from an image cache for a given image user.
*
* Returns referenced image buffer if it exists, callee is to
* call IMB_freeImBuf to de-reference the image buffer after
* it's done handling it.
*/
static ImBuf *image_get_cached_ibuf(Image *ima, ImageUser *iuser, int *r_frame, int *r_index)
{
ImBuf *ibuf = NULL;
int frame = 0, index = 0;
/* see if we already have an appropriate ibuf, with image source and type */
if (ima->source == IMA_SRC_MOVIE) {
frame = iuser ? iuser->framenr : ima->lastframe;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf_for_index_frame(ima, 0, frame);
/* XXX temp stuff? */
if (ima->lastframe != frame)
ima->tpageflag |= IMA_TPAGE_REFRESH;
ima->lastframe = frame;
}
else if (ima->source == IMA_SRC_SEQUENCE) {
if (ima->type == IMA_TYPE_IMAGE) {
frame = iuser ? iuser->framenr : ima->lastframe;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf_for_index_frame(ima, 0, frame);
2012-06-23 23:22:19 +00:00
/* XXX temp stuff? */
if (ima->lastframe != frame) {
ima->tpageflag |= IMA_TPAGE_REFRESH;
}
ima->lastframe = frame;
2012-06-23 23:22:19 +00:00
}
else if (ima->type == IMA_TYPE_MULTILAYER) {
frame = iuser ? iuser->framenr : ima->lastframe;
index = iuser ? iuser->multi_index : IMA_NO_INDEX;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf_for_index_frame(ima, index, frame);
}
}
else if (ima->source == IMA_SRC_FILE) {
if (ima->type == IMA_TYPE_IMAGE)
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf_for_index_frame(ima, IMA_NO_INDEX, 0);
else if (ima->type == IMA_TYPE_MULTILAYER)
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf_for_index_frame(ima, iuser ? iuser->multi_index : IMA_NO_INDEX, 0);
}
else if (ima->source == IMA_SRC_GENERATED) {
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf_for_index_frame(ima, IMA_NO_INDEX, 0);
2002-10-12 11:37:38 +00:00
}
else if (ima->source == IMA_SRC_VIEWER) {
/* always verify entirely, not that this shouldn't happen
* as part of texture sampling in rendering anyway, so not
* a big bottleneck */
}
if (r_frame)
*r_frame = frame;
if (r_index)
*r_index = index;
return ibuf;
}
BLI_INLINE bool image_quick_test(Image *ima, ImageUser *iuser)
{
if (ima == NULL)
return false;
if (iuser) {
if (iuser->ok == 0)
return false;
}
else if (ima->ok == 0)
return false;
return true;
}
/* Checks optional ImageUser and verifies/creates ImBuf.
*
* not thread-safe, so callee should worry about thread locks
*/
static ImBuf *image_acquire_ibuf(Image *ima, ImageUser *iuser, void **lock_r)
{
ImBuf *ibuf = NULL;
int frame = 0, index = 0;
if (lock_r)
*lock_r = NULL;
/* quick reject tests */
if (!image_quick_test(ima, iuser))
return NULL;
2012-06-23 23:22:19 +00:00
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf(ima, iuser, &frame, &index);
if (ibuf == NULL) {
/* we are sure we have to load the ibuf, using source and type */
if (ima->source == IMA_SRC_MOVIE) {
/* source is from single file, use flipbook to store ibuf */
ibuf = image_load_movie_file(ima, iuser, frame);
}
else if (ima->source == IMA_SRC_SEQUENCE) {
if (ima->type == IMA_TYPE_IMAGE) {
/* regular files, ibufs in flipbook, allows saving */
ibuf = image_load_sequence_file(ima, iuser, frame);
}
/* no else; on load the ima type can change */
if (ima->type == IMA_TYPE_MULTILAYER) {
/* only 1 layer/pass stored in imbufs, no exrhandle anim storage, no saving */
ibuf = image_load_sequence_multilayer(ima, iuser, frame);
}
}
else if (ima->source == IMA_SRC_FILE) {
if (ima->type == IMA_TYPE_IMAGE)
ibuf = image_load_image_file(ima, iuser, frame); /* cfra only for '#', this global is OK */
/* no else; on load the ima type can change */
if (ima->type == IMA_TYPE_MULTILAYER)
/* keeps render result, stores ibufs in listbase, allows saving */
ibuf = image_get_ibuf_multilayer(ima, iuser);
}
else if (ima->source == IMA_SRC_GENERATED) {
/* generated is: ibuf is allocated dynamically */
/* UV testgrid or black or solid etc */
if (ima->gen_x == 0) ima->gen_x = 1024;
if (ima->gen_y == 0) ima->gen_y = 1024;
if (ima->gen_depth == 0) ima->gen_depth = 24;
ibuf = add_ibuf_size(ima->gen_x, ima->gen_y, ima->name, ima->gen_depth, (ima->gen_flag & IMA_GEN_FLOAT) != 0, ima->gen_type,
ima->gen_color, &ima->colorspace_settings);
image_assign_ibuf(ima, ibuf, IMA_NO_INDEX, 0);
ima->ok = IMA_OK_LOADED;
}
else if (ima->source == IMA_SRC_VIEWER) {
if (ima->type == IMA_TYPE_R_RESULT) {
/* always verify entirely, and potentially
* returns pointer to release later */
ibuf = image_get_render_result(ima, iuser, lock_r);
}
else if (ima->type == IMA_TYPE_COMPOSITE) {
/* requires lock/unlock, otherwise don't return image */
if (lock_r) {
/* unlock in BKE_image_release_ibuf */
BLI_lock_thread(LOCK_VIEWER);
*lock_r = ima;
/* XXX anim play for viewer nodes not yet supported */
2013-03-31 03:28:46 +00:00
frame = 0; // XXX iuser ? iuser->framenr : 0;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf_for_index_frame(ima, 0, frame);
if (!ibuf) {
/* Composite Viewer, all handled in compositor */
/* fake ibuf, will be filled in compositor */
ibuf = IMB_allocImBuf(256, 256, 32, IB_rect);
image_assign_ibuf(ima, ibuf, 0, frame);
}
}
}
}
/* We only want movies and sequences to be memory limited. */
if (ibuf != NULL && !ELEM(ima->source, IMA_SRC_MOVIE, IMA_SRC_SEQUENCE)) {
ibuf->userflags |= IB_PERSISTENT;
}
2002-10-12 11:37:38 +00:00
}
3D view orbit option: Around Active This fixes the active object in place when orbiting the view. Choppy 15fps demo can be seen there: http://www.elysiun.com/~theeth/bf/around_active.html Image Memory Grabage Collection This adds memory handling to the image code. An image is tagged each time it is used. During a collection cycle (frequency of cycles is user defined), if an image is older than a user defined limit, its buffer gets deallocated. This also applies to gl memory buffers. Images that are loading in GL memory needs to go through two time outs before being fully deallocated: the first time out deallocated the gl memorry, the second the buffer in ram. Notes: Image buffer loaded from python gets tagged as permanent upon load. That tag is removed when python stops using the image. I might have missed some tagging spots, especially in the rendering pipeline. Someone with more knowledge about this code should check to be careful. Tagging is done on every access, for rendering, this will probably be a performance hit. A scheme should be developped to only tag when the rendering is completed. Collecting is called in draw_object, most likely not the best place to do it. Safe from undo, since using undo deallocates memory anyway (like when loading a blend file with one currently opened) Userpref DNA changes: I've changed a couple of flagging variables from short to int. Some because they needed more space, others to keep SDNA happy. Info window changes: I've grouped a couple of buttons in aligned blocks and changed the color of mutually exclusive options to make them clearer. Matt didn't do any changes on that in tuhopuu, so hopefully I'm not stepping on anyone's feet with this. Also changed double constants into floats with f in a couple of places (mostly space.c) to make compiler happier.
2005-07-06 00:33:41 +00:00
BKE_image_tag_time(ima);
return ibuf;
2002-10-12 11:37:38 +00:00
}
/* return image buffer for given image and user
*
* - will lock render result if image type is render result and lock is not NULL
* - will return NULL if image type if render or composite result and lock is NULL
*
* references the result, BKE_image_release_ibuf should be used to de-reference
*/
ImBuf *BKE_image_acquire_ibuf(Image *ima, ImageUser *iuser, void **lock_r)
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
{
ImBuf *ibuf;
BLI_spin_lock(&image_spin);
ibuf = image_acquire_ibuf(ima, iuser, lock_r);
BLI_spin_unlock(&image_spin);
return ibuf;
}
void BKE_image_release_ibuf(Image *ima, ImBuf *ibuf, void *lock)
{
if (lock) {
/* for getting image during threaded render / compositing, need to release */
if (lock == ima) {
BLI_unlock_thread(LOCK_VIEWER); /* viewer image */
}
else if (lock) {
RE_ReleaseResultImage(lock); /* render result */
BLI_unlock_thread(LOCK_VIEWER); /* view image imbuf */
}
}
if (ibuf) {
BLI_spin_lock(&image_spin);
IMB_freeImBuf(ibuf);
BLI_spin_unlock(&image_spin);
}
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
}
/* checks whether there's an image buffer for given image and user */
2014-02-03 18:55:59 +11:00
bool BKE_image_has_ibuf(Image *ima, ImageUser *iuser)
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
{
ImBuf *ibuf;
/* quick reject tests */
if (!image_quick_test(ima, iuser))
return false;
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
BLI_spin_lock(&image_spin);
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ibuf = image_get_cached_ibuf(ima, iuser, NULL, NULL);
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
if (!ibuf)
ibuf = image_acquire_ibuf(ima, iuser, NULL);
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
BLI_spin_unlock(&image_spin);
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
IMB_freeImBuf(ibuf);
return ibuf != NULL;
Render & Compositing Thread Fixes * Rendering twice or more could crash layer/pass buttons. * Compositing would crash while drawing the image. * Rendering animations could also crash drawing the image. * Compositing could crash * Starting to rendering while preview render / compo was still running could crash. * Exiting while rendering an animation would not abort the renderer properly, making Blender seemingly freeze. * Fixes theoretically possible issue with setting malloc lock with nested threads. * Drawing previews inside nodes could crash when those nodes were being rendered at the same time. There's more crashes, manipulating the scene data or undo can still crash, this commit only focuses on making sure the image buffer and render result access is thread safe. Implementation: * Rather than assuming the render result does not get freed during render, which seems to be quite difficult to do given that e.g. the compositor is allowed to change the size of the buffer or output different passes, the render result is now protected with a read/write mutex. * The read/write mutex allows multiple readers (and pixel writers) at the same time, but only allows one writer to manipulate the data structure. * Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access images being rendered, cases where this is not needed (most code) can still use BKE_image_get_ibuf. * The job manager now allows only one rendering job at the same time, rather than the G.rendering check which was not reliable.
2009-09-30 18:18:32 +00:00
}
2002-10-12 11:37:38 +00:00
/* ******** Pool for image buffers ******** */
typedef struct ImagePoolEntry {
struct ImagePoolEntry *next, *prev;
Image *image;
ImBuf *ibuf;
int index;
int frame;
} ImagePoolEntry;
typedef struct ImagePool {
ListBase image_buffers;
} ImagePool;
ImagePool *BKE_image_pool_new(void)
{
ImagePool *pool = MEM_callocN(sizeof(ImagePool), "Image Pool");
return pool;
}
void BKE_image_pool_free(ImagePool *pool)
{
ImagePoolEntry *entry, *next_entry;
/* use single lock to dereference all the image buffers */
BLI_spin_lock(&image_spin);
for (entry = pool->image_buffers.first; entry; entry = next_entry) {
next_entry = entry->next;
if (entry->ibuf)
IMB_freeImBuf(entry->ibuf);
MEM_freeN(entry);
}
BLI_spin_unlock(&image_spin);
MEM_freeN(pool);
}
BLI_INLINE ImBuf *image_pool_find_entry(ImagePool *pool, Image *image, int frame, int index, bool *found)
{
ImagePoolEntry *entry;
*found = false;
for (entry = pool->image_buffers.first; entry; entry = entry->next) {
if (entry->image == image && entry->frame == frame && entry->index == index) {
*found = true;
return entry->ibuf;
}
}
return NULL;
}
ImBuf *BKE_image_pool_acquire_ibuf(Image *ima, ImageUser *iuser, ImagePool *pool)
{
ImBuf *ibuf;
int index, frame;
bool found;
if (!image_quick_test(ima, iuser))
return NULL;
if (pool == NULL) {
/* pool could be NULL, in this case use general acquire function */
return BKE_image_acquire_ibuf(ima, iuser, NULL);
}
image_get_frame_and_index(ima, iuser, &frame, &index);
ibuf = image_pool_find_entry(pool, ima, frame, index, &found);
if (found)
return ibuf;
BLI_spin_lock(&image_spin);
ibuf = image_pool_find_entry(pool, ima, frame, index, &found);
/* will also create entry even in cases image buffer failed to load,
* prevents trying to load the same buggy file multiple times
*/
if (!found) {
ImagePoolEntry *entry;
ibuf = image_acquire_ibuf(ima, iuser, NULL);
entry = MEM_callocN(sizeof(ImagePoolEntry), "Image Pool Entry");
entry->image = ima;
entry->frame = frame;
entry->index = index;
entry->ibuf = ibuf;
BLI_addtail(&pool->image_buffers, entry);
}
BLI_spin_unlock(&image_spin);
return ibuf;
}
void BKE_image_pool_release_ibuf(Image *ima, ImBuf *ibuf, ImagePool *pool)
{
/* if pool wasn't actually used, use general release stuff,
* for pools image buffers will be dereferenced on pool free
*/
if (pool == NULL) {
BKE_image_release_ibuf(ima, ibuf, NULL);
}
}
2014-02-03 18:55:59 +11:00
int BKE_image_user_frame_get(const ImageUser *iuser, int cfra, int fieldnr, bool *r_is_in_range)
{
const int len = (iuser->fie_ima * iuser->frames) / 2;
if (r_is_in_range) {
*r_is_in_range = false;
}
if (len == 0) {
return 0;
}
else {
int framenr;
cfra = cfra - iuser->sfra + 1;
/* cyclic */
if (iuser->cycl) {
cfra = ((cfra) % len);
if (cfra < 0) cfra += len;
if (cfra == 0) cfra = len;
if (r_is_in_range) {
*r_is_in_range = true;
}
}
if (cfra < 0) {
cfra = 0;
}
else if (cfra > len) {
cfra = len;
}
else {
if (r_is_in_range) {
*r_is_in_range = true;
}
}
/* convert current frame to current field */
cfra = 2 * (cfra);
if (fieldnr) cfra++;
/* transform to images space */
framenr = (cfra + iuser->fie_ima - 2) / iuser->fie_ima;
if (framenr > iuser->frames) framenr = iuser->frames;
if (iuser->cycl) {
framenr = ((framenr) % len);
while (framenr < 0) framenr += len;
if (framenr == 0) framenr = len;
}
/* important to apply after else we cant loop on frames 100 - 110 for eg. */
framenr += iuser->offset;
return framenr;
}
}
void BKE_image_user_frame_calc(ImageUser *iuser, int cfra, int fieldnr)
{
if (iuser) {
2014-02-03 18:55:59 +11:00
bool is_in_range;
const int framenr = BKE_image_user_frame_get(iuser, cfra, fieldnr, &is_in_range);
if (is_in_range) {
iuser->flag |= IMA_USER_FRAME_IN_RANGE;
}
else {
iuser->flag &= ~IMA_USER_FRAME_IN_RANGE;
}
2012-06-23 23:22:19 +00:00
/* allows image users to handle redraws */
if (iuser->flag & IMA_ANIM_ALWAYS)
if (framenr != iuser->framenr)
iuser->flag |= IMA_ANIM_REFRESHED;
2012-06-23 23:22:19 +00:00
iuser->framenr = framenr;
if (iuser->ok == 0) iuser->ok = 1;
}
}
TexFace to Material Settings big patch Summary: ======== The idea here is to move the texface options into the material panel. For images with the change please visit: http://code.blender.org/index.php/2011/09/bge-material-texface-changes 1 - Some of the legacy problems 2.49 and 2.5x has with the texface system: ========================================================================== 1.1) Shadow, Bilboard and Halo are mutual exclusive (in the code), yet you can select a face to be more than one mode. 1.2) Sort only works for blend Alpha yet it's an option regardless of the Transparency Blend you pick. 1.3) Shared doesn't affect anything in BGE. 1.4) ObColor only works for Text objects (old bitmap texts) when using Texture Face Materials. (not address yet, I so far ignored obcolor) 2 - Notes: ============ 2.1) Now "Use Face Textures" in material Option panel will work in Multitexture even if there is no texture channel. 2.2) In FaceTexture mode it will use TexFace all the time, even if you don't check the "Use Texture Face" option in the UI. It's a matter of decision, since the code for either way is there. I decided by the solution that makes the creation of a material fast - in this mode the user doesn't need to mess with textures or this "Use Texture Face" option at all. I'm not strong in my opinion here. But I think if we don't have this then what is the point of the Texture Face mode? 2.3) I kept references for tface only when we need the image, UV or the tiling setting. It should help later when/if we split the Image and UV layers from the tface struct (Campbell and Brecht proposal). 3 - Changes in a Nutshell: ========================== 3.1) "Texture Face" panel (in the Mesh/Object Data panel) no longer exists. Those settings are all part of the material properties, visible when Game Render is set. 3.2) "Texture Face" Shading mode (in the Render panel) is now called “Single Texture”, it needs a material for special settings (e.g. Billboard, Alpha Sort, …). 3.3) New options in the Material Panel * Shadeless option in the Material panel is now supported for all three Shading modes. * Physics is now toggleable, this is the old Collision option. * Two Side (on) is now called Back Culling (off). * Alpha Sort is one of the Alpha options, together (and mutually exclusive) to Alpha Blend, Alpha Clip, Add and Opaque (i.e. solid). * Shadow, Billboard and Halo are grouped in the “Face Orientation” property. * "Face Textures" and "Face Textures Alpha" (under Options) can be used for all but GLSL shading mode (to be supported in GLSL eventually). * The backend in the game engine is still the same as before. The only changes are in the interface and in the way you need to think your materials. The bottomline is: It’s no longer possible to share materials between faces that do not share the same game properties. 4 - Acknowledgment: ================== Mike Pan for the design discussions, and testing along the whole development process. Vitor Balbio for the first hands-on code with the interface changes. That helped me a lot to push me into work on that. Benoit Bolsee and Brecht van Lommel for patch review (* no one reviewed the whole patch, or the latest iteractions, so I still hold liability for any problems). Blender artists that gave feedback and helped testing the patch. Patch review and original documentation can be found here: http://wiki.blender.org/index.php/User:Dfelinto/TexFace http://codereview.appspot.com/4289041/
2011-09-19 19:55:59 +00:00
void BKE_image_user_check_frame_calc(ImageUser *iuser, int cfra, int fieldnr)
{
if ((iuser->flag & IMA_ANIM_ALWAYS) || (iuser->flag & IMA_NEED_FRAME_RECALC)) {
BKE_image_user_frame_calc(iuser, cfra, fieldnr);
iuser->flag &= ~IMA_NEED_FRAME_RECALC;
}
}
/* goes over all ImageUsers, and sets frame numbers if auto-refresh is set */
static void image_update_frame(struct Image *UNUSED(ima), struct ImageUser *iuser, void *customdata)
{
int cfra = *(int *)customdata;
BKE_image_user_check_frame_calc(iuser, cfra, 0);
}
void BKE_image_update_frame(const Main *bmain, int cfra)
{
BKE_image_walk_all_users(bmain, &cfra, image_update_frame);
}
void BKE_image_user_file_path(ImageUser *iuser, Image *ima, char *filepath)
{
BLI_strncpy(filepath, ima->name, FILE_MAX);
if (ima->source == IMA_SRC_SEQUENCE) {
char head[FILE_MAX], tail[FILE_MAX];
unsigned short numlen;
int frame = iuser ? iuser->framenr : ima->lastframe;
BLI_stringdec(filepath, head, tail, &numlen);
BLI_stringenc(filepath, head, tail, numlen, frame);
}
BLI_path_abs(filepath, ID_BLEND_PATH(G.main, &ima->id));
}
2014-02-03 18:55:59 +11:00
bool BKE_image_has_alpha(struct Image *image)
TexFace to Material Settings big patch Summary: ======== The idea here is to move the texface options into the material panel. For images with the change please visit: http://code.blender.org/index.php/2011/09/bge-material-texface-changes 1 - Some of the legacy problems 2.49 and 2.5x has with the texface system: ========================================================================== 1.1) Shadow, Bilboard and Halo are mutual exclusive (in the code), yet you can select a face to be more than one mode. 1.2) Sort only works for blend Alpha yet it's an option regardless of the Transparency Blend you pick. 1.3) Shared doesn't affect anything in BGE. 1.4) ObColor only works for Text objects (old bitmap texts) when using Texture Face Materials. (not address yet, I so far ignored obcolor) 2 - Notes: ============ 2.1) Now "Use Face Textures" in material Option panel will work in Multitexture even if there is no texture channel. 2.2) In FaceTexture mode it will use TexFace all the time, even if you don't check the "Use Texture Face" option in the UI. It's a matter of decision, since the code for either way is there. I decided by the solution that makes the creation of a material fast - in this mode the user doesn't need to mess with textures or this "Use Texture Face" option at all. I'm not strong in my opinion here. But I think if we don't have this then what is the point of the Texture Face mode? 2.3) I kept references for tface only when we need the image, UV or the tiling setting. It should help later when/if we split the Image and UV layers from the tface struct (Campbell and Brecht proposal). 3 - Changes in a Nutshell: ========================== 3.1) "Texture Face" panel (in the Mesh/Object Data panel) no longer exists. Those settings are all part of the material properties, visible when Game Render is set. 3.2) "Texture Face" Shading mode (in the Render panel) is now called “Single Texture”, it needs a material for special settings (e.g. Billboard, Alpha Sort, …). 3.3) New options in the Material Panel * Shadeless option in the Material panel is now supported for all three Shading modes. * Physics is now toggleable, this is the old Collision option. * Two Side (on) is now called Back Culling (off). * Alpha Sort is one of the Alpha options, together (and mutually exclusive) to Alpha Blend, Alpha Clip, Add and Opaque (i.e. solid). * Shadow, Billboard and Halo are grouped in the “Face Orientation” property. * "Face Textures" and "Face Textures Alpha" (under Options) can be used for all but GLSL shading mode (to be supported in GLSL eventually). * The backend in the game engine is still the same as before. The only changes are in the interface and in the way you need to think your materials. The bottomline is: It’s no longer possible to share materials between faces that do not share the same game properties. 4 - Acknowledgment: ================== Mike Pan for the design discussions, and testing along the whole development process. Vitor Balbio for the first hands-on code with the interface changes. That helped me a lot to push me into work on that. Benoit Bolsee and Brecht van Lommel for patch review (* no one reviewed the whole patch, or the latest iteractions, so I still hold liability for any problems). Blender artists that gave feedback and helped testing the patch. Patch review and original documentation can be found here: http://wiki.blender.org/index.php/User:Dfelinto/TexFace http://codereview.appspot.com/4289041/
2011-09-19 19:55:59 +00:00
{
ImBuf *ibuf;
void *lock;
int planes;
2012-06-23 23:22:19 +00:00
ibuf = BKE_image_acquire_ibuf(image, NULL, &lock);
planes = (ibuf ? ibuf->planes : 0);
BKE_image_release_ibuf(image, ibuf, lock);
TexFace to Material Settings big patch Summary: ======== The idea here is to move the texface options into the material panel. For images with the change please visit: http://code.blender.org/index.php/2011/09/bge-material-texface-changes 1 - Some of the legacy problems 2.49 and 2.5x has with the texface system: ========================================================================== 1.1) Shadow, Bilboard and Halo are mutual exclusive (in the code), yet you can select a face to be more than one mode. 1.2) Sort only works for blend Alpha yet it's an option regardless of the Transparency Blend you pick. 1.3) Shared doesn't affect anything in BGE. 1.4) ObColor only works for Text objects (old bitmap texts) when using Texture Face Materials. (not address yet, I so far ignored obcolor) 2 - Notes: ============ 2.1) Now "Use Face Textures" in material Option panel will work in Multitexture even if there is no texture channel. 2.2) In FaceTexture mode it will use TexFace all the time, even if you don't check the "Use Texture Face" option in the UI. It's a matter of decision, since the code for either way is there. I decided by the solution that makes the creation of a material fast - in this mode the user doesn't need to mess with textures or this "Use Texture Face" option at all. I'm not strong in my opinion here. But I think if we don't have this then what is the point of the Texture Face mode? 2.3) I kept references for tface only when we need the image, UV or the tiling setting. It should help later when/if we split the Image and UV layers from the tface struct (Campbell and Brecht proposal). 3 - Changes in a Nutshell: ========================== 3.1) "Texture Face" panel (in the Mesh/Object Data panel) no longer exists. Those settings are all part of the material properties, visible when Game Render is set. 3.2) "Texture Face" Shading mode (in the Render panel) is now called “Single Texture”, it needs a material for special settings (e.g. Billboard, Alpha Sort, …). 3.3) New options in the Material Panel * Shadeless option in the Material panel is now supported for all three Shading modes. * Physics is now toggleable, this is the old Collision option. * Two Side (on) is now called Back Culling (off). * Alpha Sort is one of the Alpha options, together (and mutually exclusive) to Alpha Blend, Alpha Clip, Add and Opaque (i.e. solid). * Shadow, Billboard and Halo are grouped in the “Face Orientation” property. * "Face Textures" and "Face Textures Alpha" (under Options) can be used for all but GLSL shading mode (to be supported in GLSL eventually). * The backend in the game engine is still the same as before. The only changes are in the interface and in the way you need to think your materials. The bottomline is: It’s no longer possible to share materials between faces that do not share the same game properties. 4 - Acknowledgment: ================== Mike Pan for the design discussions, and testing along the whole development process. Vitor Balbio for the first hands-on code with the interface changes. That helped me a lot to push me into work on that. Benoit Bolsee and Brecht van Lommel for patch review (* no one reviewed the whole patch, or the latest iteractions, so I still hold liability for any problems). Blender artists that gave feedback and helped testing the patch. Patch review and original documentation can be found here: http://wiki.blender.org/index.php/User:Dfelinto/TexFace http://codereview.appspot.com/4289041/
2011-09-19 19:55:59 +00:00
if (planes == 32)
2014-12-01 17:11:18 +01:00
return true;
TexFace to Material Settings big patch Summary: ======== The idea here is to move the texface options into the material panel. For images with the change please visit: http://code.blender.org/index.php/2011/09/bge-material-texface-changes 1 - Some of the legacy problems 2.49 and 2.5x has with the texface system: ========================================================================== 1.1) Shadow, Bilboard and Halo are mutual exclusive (in the code), yet you can select a face to be more than one mode. 1.2) Sort only works for blend Alpha yet it's an option regardless of the Transparency Blend you pick. 1.3) Shared doesn't affect anything in BGE. 1.4) ObColor only works for Text objects (old bitmap texts) when using Texture Face Materials. (not address yet, I so far ignored obcolor) 2 - Notes: ============ 2.1) Now "Use Face Textures" in material Option panel will work in Multitexture even if there is no texture channel. 2.2) In FaceTexture mode it will use TexFace all the time, even if you don't check the "Use Texture Face" option in the UI. It's a matter of decision, since the code for either way is there. I decided by the solution that makes the creation of a material fast - in this mode the user doesn't need to mess with textures or this "Use Texture Face" option at all. I'm not strong in my opinion here. But I think if we don't have this then what is the point of the Texture Face mode? 2.3) I kept references for tface only when we need the image, UV or the tiling setting. It should help later when/if we split the Image and UV layers from the tface struct (Campbell and Brecht proposal). 3 - Changes in a Nutshell: ========================== 3.1) "Texture Face" panel (in the Mesh/Object Data panel) no longer exists. Those settings are all part of the material properties, visible when Game Render is set. 3.2) "Texture Face" Shading mode (in the Render panel) is now called “Single Texture”, it needs a material for special settings (e.g. Billboard, Alpha Sort, …). 3.3) New options in the Material Panel * Shadeless option in the Material panel is now supported for all three Shading modes. * Physics is now toggleable, this is the old Collision option. * Two Side (on) is now called Back Culling (off). * Alpha Sort is one of the Alpha options, together (and mutually exclusive) to Alpha Blend, Alpha Clip, Add and Opaque (i.e. solid). * Shadow, Billboard and Halo are grouped in the “Face Orientation” property. * "Face Textures" and "Face Textures Alpha" (under Options) can be used for all but GLSL shading mode (to be supported in GLSL eventually). * The backend in the game engine is still the same as before. The only changes are in the interface and in the way you need to think your materials. The bottomline is: It’s no longer possible to share materials between faces that do not share the same game properties. 4 - Acknowledgment: ================== Mike Pan for the design discussions, and testing along the whole development process. Vitor Balbio for the first hands-on code with the interface changes. That helped me a lot to push me into work on that. Benoit Bolsee and Brecht van Lommel for patch review (* no one reviewed the whole patch, or the latest iteractions, so I still hold liability for any problems). Blender artists that gave feedback and helped testing the patch. Patch review and original documentation can be found here: http://wiki.blender.org/index.php/User:Dfelinto/TexFace http://codereview.appspot.com/4289041/
2011-09-19 19:55:59 +00:00
else
2014-12-01 17:11:18 +01:00
return false;
TexFace to Material Settings big patch Summary: ======== The idea here is to move the texface options into the material panel. For images with the change please visit: http://code.blender.org/index.php/2011/09/bge-material-texface-changes 1 - Some of the legacy problems 2.49 and 2.5x has with the texface system: ========================================================================== 1.1) Shadow, Bilboard and Halo are mutual exclusive (in the code), yet you can select a face to be more than one mode. 1.2) Sort only works for blend Alpha yet it's an option regardless of the Transparency Blend you pick. 1.3) Shared doesn't affect anything in BGE. 1.4) ObColor only works for Text objects (old bitmap texts) when using Texture Face Materials. (not address yet, I so far ignored obcolor) 2 - Notes: ============ 2.1) Now "Use Face Textures" in material Option panel will work in Multitexture even if there is no texture channel. 2.2) In FaceTexture mode it will use TexFace all the time, even if you don't check the "Use Texture Face" option in the UI. It's a matter of decision, since the code for either way is there. I decided by the solution that makes the creation of a material fast - in this mode the user doesn't need to mess with textures or this "Use Texture Face" option at all. I'm not strong in my opinion here. But I think if we don't have this then what is the point of the Texture Face mode? 2.3) I kept references for tface only when we need the image, UV or the tiling setting. It should help later when/if we split the Image and UV layers from the tface struct (Campbell and Brecht proposal). 3 - Changes in a Nutshell: ========================== 3.1) "Texture Face" panel (in the Mesh/Object Data panel) no longer exists. Those settings are all part of the material properties, visible when Game Render is set. 3.2) "Texture Face" Shading mode (in the Render panel) is now called “Single Texture”, it needs a material for special settings (e.g. Billboard, Alpha Sort, …). 3.3) New options in the Material Panel * Shadeless option in the Material panel is now supported for all three Shading modes. * Physics is now toggleable, this is the old Collision option. * Two Side (on) is now called Back Culling (off). * Alpha Sort is one of the Alpha options, together (and mutually exclusive) to Alpha Blend, Alpha Clip, Add and Opaque (i.e. solid). * Shadow, Billboard and Halo are grouped in the “Face Orientation” property. * "Face Textures" and "Face Textures Alpha" (under Options) can be used for all but GLSL shading mode (to be supported in GLSL eventually). * The backend in the game engine is still the same as before. The only changes are in the interface and in the way you need to think your materials. The bottomline is: It’s no longer possible to share materials between faces that do not share the same game properties. 4 - Acknowledgment: ================== Mike Pan for the design discussions, and testing along the whole development process. Vitor Balbio for the first hands-on code with the interface changes. That helped me a lot to push me into work on that. Benoit Bolsee and Brecht van Lommel for patch review (* no one reviewed the whole patch, or the latest iteractions, so I still hold liability for any problems). Blender artists that gave feedback and helped testing the patch. Patch review and original documentation can be found here: http://wiki.blender.org/index.php/User:Dfelinto/TexFace http://codereview.appspot.com/4289041/
2011-09-19 19:55:59 +00:00
}
void BKE_image_get_size(Image *image, ImageUser *iuser, int *width, int *height)
{
ImBuf *ibuf = NULL;
void *lock;
ibuf = BKE_image_acquire_ibuf(image, iuser, &lock);
if (ibuf && ibuf->x > 0 && ibuf->y > 0) {
*width = ibuf->x;
*height = ibuf->y;
}
else {
*width = IMG_SIZE_FALLBACK;
*height = IMG_SIZE_FALLBACK;
}
BKE_image_release_ibuf(image, ibuf, lock);
}
void BKE_image_get_size_fl(Image *image, ImageUser *iuser, float size[2])
{
int width, height;
BKE_image_get_size(image, iuser, &width, &height);
size[0] = (float)width;
size[1] = (float)height;
}
void BKE_image_get_aspect(Image *image, float *aspx, float *aspy)
{
*aspx = 1.0;
/* x is always 1 */
if (image)
*aspy = image->aspy / image->aspx;
else
*aspy = 1.0f;
}
unsigned char *BKE_image_get_pixels_for_frame(struct Image *image, int frame)
{
ImageUser iuser = {NULL};
void *lock;
ImBuf *ibuf;
unsigned char *pixels = NULL;
iuser.framenr = frame;
iuser.ok = true;
ibuf = BKE_image_acquire_ibuf(image, &iuser, &lock);
if (ibuf) {
pixels = (unsigned char *) ibuf->rect;
if (pixels)
pixels = MEM_dupallocN(pixels);
BKE_image_release_ibuf(image, ibuf, lock);
}
if (!pixels)
return NULL;
return pixels;
}
float *BKE_image_get_float_pixels_for_frame(struct Image *image, int frame)
{
ImageUser iuser = {NULL};
void *lock;
ImBuf *ibuf;
float *pixels = NULL;
iuser.framenr = frame;
iuser.ok = true;
ibuf = BKE_image_acquire_ibuf(image, &iuser, &lock);
if (ibuf) {
pixels = ibuf->rect_float;
if (pixels)
pixels = MEM_dupallocN(pixels);
BKE_image_release_ibuf(image, ibuf, lock);
}
if (!pixels)
return NULL;
return pixels;
}
int BKE_image_sequence_guess_offset(Image *image)
{
unsigned short numlen;
char head[FILE_MAX], tail[FILE_MAX];
char num[FILE_MAX] = {0};
BLI_stringdec(image->name, head, tail, &numlen);
BLI_strncpy(num, image->name + strlen(head), numlen + 1);
return atoi(num);
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
/**
* Checks the image buffer changes (not keyframed values)
*
* to see if we need to call #BKE_image_user_check_frame_calc
*/
bool BKE_image_is_animated(Image *image)
{
return ELEM(image->source, IMA_SRC_MOVIE, IMA_SRC_SEQUENCE);
}
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
bool BKE_image_is_dirty(Image *image)
{
bool is_dirty = false;
BLI_spin_lock(&image_spin);
if (image->cache != NULL) {
struct MovieCacheIter *iter = IMB_moviecacheIter_new(image->cache);
while (!IMB_moviecacheIter_done(iter)) {
ImBuf *ibuf = IMB_moviecacheIter_getImBuf(iter);
if (ibuf->userflags & IB_BITMAPDIRTY) {
is_dirty = true;
break;
}
IMB_moviecacheIter_step(iter);
}
IMB_moviecacheIter_free(iter);
}
BLI_spin_unlock(&image_spin);
return is_dirty;
}
void BKE_image_file_format_set(Image *image, int ftype)
{
#if 0
ImBuf *ibuf = BKE_image_acquire_ibuf(image, NULL, NULL);
if (ibuf) {
ibuf->ftype = ftype;
}
BKE_image_release_ibuf(image, ibuf, NULL);
#endif
BLI_spin_lock(&image_spin);
if (image->cache != NULL) {
struct MovieCacheIter *iter = IMB_moviecacheIter_new(image->cache);
while (!IMB_moviecacheIter_done(iter)) {
ImBuf *ibuf = IMB_moviecacheIter_getImBuf(iter);
ibuf->ftype = ftype;
IMB_moviecacheIter_step(iter);
}
IMB_moviecacheIter_free(iter);
}
BLI_spin_unlock(&image_spin);
}
bool BKE_image_has_loaded_ibuf(Image *image)
{
bool has_loaded_ibuf = false;
BLI_spin_lock(&image_spin);
if (image->cache != NULL) {
struct MovieCacheIter *iter = IMB_moviecacheIter_new(image->cache);
while (!IMB_moviecacheIter_done(iter)) {
has_loaded_ibuf = true;
break;
}
IMB_moviecacheIter_free(iter);
}
BLI_spin_unlock(&image_spin);
return has_loaded_ibuf;
}
/* References the result, BKE_image_release_ibuf is to be called to de-reference.
* Use lock=NULL when calling BKE_image_release_ibuf().
*/
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
ImBuf *BKE_image_get_ibuf_with_name(Image *image, const char *name)
{
ImBuf *ibuf = NULL;
BLI_spin_lock(&image_spin);
if (image->cache != NULL) {
struct MovieCacheIter *iter = IMB_moviecacheIter_new(image->cache);
while (!IMB_moviecacheIter_done(iter)) {
ImBuf *current_ibuf = IMB_moviecacheIter_getImBuf(iter);
if (STREQ(current_ibuf->name, name)) {
ibuf = current_ibuf;
IMB_refImBuf(ibuf);
break;
}
IMB_moviecacheIter_step(iter);
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
}
IMB_moviecacheIter_free(iter);
}
BLI_spin_unlock(&image_spin);
return ibuf;
}
/* References the result, BKE_image_release_ibuf is to be called to de-reference.
* Use lock=NULL when calling BKE_image_release_ibuf().
Image cache rewrite to using generic movie cache Summary: Behaves very much the same as cache for Movie Clip datablock: - Image now have `MovieCache *cache` field which replaced legacy `ListBase ibufs`. This allows image datablock to easily keep of image buffers which are owned by itself. This field isn't saved to the file and getting restored on undo steps. However, cache limit is global for movies, sequences and image datablocks now. So overall cached image buffers size will not go above cache limit size in user preferences. - Image buffers which are marked as BITMAPDIRTY will never be freed from the cache. - Added utility function to iterate over image buffers saved in movie cache. - Movie cache cleanup check callback now have ImBuf argument which can be used in a condition of cleanup. - Added some utility functions which replaces legacy ibufs iterations with image cache iteration which happens from inside a lock. - Fixed `image_mem_size()` which was only counting one of the buffers if both float and byte buffer present. Additional notes: - `BKE_image_get_first_ibuf()` is rather stupid, but direct access to ibufs->first was also the same stupid idea. Would consider avoid this function is another project. - There are some places which doesn't look threadsafe, but they already were not so much threadsafe anyway before. So think not a big deal with solving this later. Finally solves infinite memory usage by image sequences! :) Reviewers: brecht, campbellbarton Reviewed By: brecht CC: sebastian_k Differential Revision: http://developer.blender.org/D95
2013-12-13 16:22:08 +06:00
*
* TODO(sergey): This is actually "get first entry from the cache", which is
* not so much predictable. But using first loaded image buffer
* was also malicious logic and all the areas which uses this
* function are to be re-considered.
*/
ImBuf *BKE_image_get_first_ibuf(Image *image)
{
ImBuf *ibuf = NULL;
BLI_spin_lock(&image_spin);
if (image->cache != NULL) {
struct MovieCacheIter *iter = IMB_moviecacheIter_new(image->cache);
while (!IMB_moviecacheIter_done(iter)) {
ibuf = IMB_moviecacheIter_getImBuf(iter);
IMB_refImBuf(ibuf);
break;
}
IMB_moviecacheIter_free(iter);
}
BLI_spin_unlock(&image_spin);
return ibuf;
}