This repository has been archived on 2023-10-09.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/editors/render/render_internal.c

1715 lines
49 KiB
C
Raw Normal View History

/*
* ***** BEGIN GPL LICENSE BLOCK *****
*
* 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.
*
* 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,
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*
* The Original Code is Copyright (C) 2008 Blender Foundation.
* All rights reserved.
*
*
* ***** END GPL LICENSE BLOCK *****
*/
2011-02-27 20:29:51 +00:00
/** \file blender/editors/render/render_internal.c
* \ingroup edrend
*/
#include <math.h>
#include <string.h>
#include <stddef.h>
#include "MEM_guardedalloc.h"
#include "BLI_listbase.h"
#include "BLI_rect.h"
#include "BLI_timecode.h"
#include "BLI_math.h"
#include "BLI_threads.h"
#include "BLI_utildefines.h"
#include "PIL_time.h"
#include "BLT_translation.h"
#include "DNA_object_types.h"
#include "DNA_scene_types.h"
#include "DNA_view3d_types.h"
#include "DNA_userdef_types.h"
#include "BKE_blender_undo.h"
#include "BKE_blender_version.h"
Multi-View and Stereo 3D Official Documentation: http://www.blender.org/manual/render/workflows/multiview.html Implemented Features ==================== Builtin Stereo Camera * Convergence Mode * Interocular Distance * Convergence Distance * Pivot Mode Viewport * Cameras * Plane * Volume Compositor * View Switch Node * Image Node Multi-View OpenEXR support Sequencer * Image/Movie Strips 'Use Multiview' UV/Image Editor * Option to see Multi-View images in Stereo-3D or its individual images * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images I/O * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images Scene Render Views * Ability to have an arbitrary number of views in the scene Missing Bits ============ First rule of Multi-View bug report: If something is not working as it should *when Views is off* this is a severe bug, do mention this in the report. Second rule is, if something works *when Views is off* but doesn't (or crashes) when *Views is on*, this is a important bug. Do mention this in the report. Everything else is likely small todos, and may wait until we are sure none of the above is happening. Apart from that there are those known issues: * Compositor Image Node poorly working for Multi-View OpenEXR (this was working prefectly before the 'Use Multi-View' functionality) * Selecting camera from Multi-View when looking from camera is problematic * Animation Playback (ctrl+F11) doesn't support stereo formats * Wrong filepath when trying to play back animated scene * Viewport Rendering doesn't support Multi-View * Overscan Rendering * Fullscreen display modes need to warn the user * Object copy should be aware of views suffix Acknowledgments =============== * Francesco Siddi for the help with the original feature specs and design * Brecht Van Lommel for the original review of the code and design early on * Blender Foundation for the Development Fund to support the project wrap up Final patch reviewers: * Antony Riakiotakis (psy-fi) * Campbell Barton (ideasman42) * Julian Eisel (Severin) * Sergey Sharybin (nazgul) * Thomas Dinged (dingto) Code contributors of the original branch in github: * Alexey Akishin * Gabriel Caraballo
2015-04-06 10:40:12 -03:00
#include "BKE_camera.h"
#include "BKE_context.h"
#include "BKE_colortools.h"
#include "BKE_depsgraph.h"
#include "BKE_global.h"
#include "BKE_image.h"
#include "BKE_library.h"
#include "BKE_main.h"
#include "BKE_node.h"
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
#include "BKE_object.h"
#include "BKE_report.h"
#include "BKE_sequencer.h"
#include "BKE_screen.h"
#include "BKE_scene.h"
#include "BKE_undo_system.h"
#include "WM_api.h"
#include "WM_types.h"
#include "ED_object.h"
#include "ED_render.h"
#include "ED_screen.h"
#include "ED_util.h"
#include "ED_view3d.h"
#include "RE_pipeline.h"
#include "RE_engine.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"
#include "IMB_imbuf_types.h"
#include "BIF_gl.h"
#include "BIF_glutil.h"
#include "RNA_access.h"
#include "RNA_define.h"
#include "BLO_undofile.h"
#include "render_intern.h"
/* Render Callbacks */
static int render_break(void *rjv);
typedef struct RenderJob {
Main *main;
Scene *scene;
Scene *current_scene;
Render *re;
SceneRenderLayer *srl;
struct Object *camera_override;
int lay_override;
bool v3d_override;
bool anim, write_still;
Image *image;
ImageUser iuser;
bool image_outdated;
short *stop;
short *do_update;
float *progress;
ReportList *reports;
int orig_layer;
int last_layer;
ScrArea *sa;
ColorManagedViewSettings view_settings;
ColorManagedDisplaySettings display_settings;
bool supports_glsl_draw;
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
bool interface_locked;
} RenderJob;
/* called inside thread! */
Multi-View and Stereo 3D Official Documentation: http://www.blender.org/manual/render/workflows/multiview.html Implemented Features ==================== Builtin Stereo Camera * Convergence Mode * Interocular Distance * Convergence Distance * Pivot Mode Viewport * Cameras * Plane * Volume Compositor * View Switch Node * Image Node Multi-View OpenEXR support Sequencer * Image/Movie Strips 'Use Multiview' UV/Image Editor * Option to see Multi-View images in Stereo-3D or its individual images * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images I/O * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images Scene Render Views * Ability to have an arbitrary number of views in the scene Missing Bits ============ First rule of Multi-View bug report: If something is not working as it should *when Views is off* this is a severe bug, do mention this in the report. Second rule is, if something works *when Views is off* but doesn't (or crashes) when *Views is on*, this is a important bug. Do mention this in the report. Everything else is likely small todos, and may wait until we are sure none of the above is happening. Apart from that there are those known issues: * Compositor Image Node poorly working for Multi-View OpenEXR (this was working prefectly before the 'Use Multi-View' functionality) * Selecting camera from Multi-View when looking from camera is problematic * Animation Playback (ctrl+F11) doesn't support stereo formats * Wrong filepath when trying to play back animated scene * Viewport Rendering doesn't support Multi-View * Overscan Rendering * Fullscreen display modes need to warn the user * Object copy should be aware of views suffix Acknowledgments =============== * Francesco Siddi for the help with the original feature specs and design * Brecht Van Lommel for the original review of the code and design early on * Blender Foundation for the Development Fund to support the project wrap up Final patch reviewers: * Antony Riakiotakis (psy-fi) * Campbell Barton (ideasman42) * Julian Eisel (Severin) * Sergey Sharybin (nazgul) * Thomas Dinged (dingto) Code contributors of the original branch in github: * Alexey Akishin * Gabriel Caraballo
2015-04-06 10:40:12 -03:00
static void image_buffer_rect_update(RenderJob *rj, RenderResult *rr, ImBuf *ibuf, ImageUser *iuser, volatile rcti *renrect, const char *viewname)
{
Scene *scene = rj->scene;
const float *rectf = NULL;
int ymin, ymax, xmin, xmax;
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
int rymin, rxmin;
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
int linear_stride, linear_offset_x, linear_offset_y;
ColorManagedViewSettings *view_settings;
ColorManagedDisplaySettings *display_settings;
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
/* Exception for exr tiles -- display buffer conversion happens here,
* NOT in the color management pipeline.
*/
if (ibuf->userflags & IB_DISPLAY_BUFFER_INVALID &&
rr->do_exr_tile == false)
{
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
/* The whole image buffer it so be color managed again anyway. */
return;
}
/* if renrect argument, we only refresh scanlines */
if (renrect) {
/* if (ymax == recty), rendering of layer is ready, we should not draw, other things happen... */
2012-03-29 22:42:32 +00:00
if (rr->renlay == NULL || renrect->ymax >= rr->recty)
return;
/* xmin here is first subrect x coord, xmax defines subrect width */
xmin = renrect->xmin + rr->crop;
xmax = renrect->xmax - xmin + rr->crop;
2012-03-29 22:42:32 +00:00
if (xmax < 2)
return;
2012-03-29 22:42:32 +00:00
ymin = renrect->ymin + rr->crop;
ymax = renrect->ymax - ymin + rr->crop;
if (ymax < 2)
return;
renrect->ymin = renrect->ymax;
}
else {
xmin = ymin = rr->crop;
2012-03-29 22:42:32 +00:00
xmax = rr->rectx - 2 * rr->crop;
ymax = rr->recty - 2 * rr->crop;
}
/* xmin ymin is in tile coords. transform to ibuf */
2012-03-29 22:42:32 +00:00
rxmin = rr->tilerect.xmin + xmin;
if (rxmin >= ibuf->x) return;
2012-03-29 22:42:32 +00:00
rymin = rr->tilerect.ymin + ymin;
if (rymin >= ibuf->y) return;
if (rxmin + xmax > ibuf->x)
2012-03-29 22:42:32 +00:00
xmax = ibuf->x - rxmin;
if (rymin + ymax > ibuf->y)
2012-03-29 22:42:32 +00:00
ymax = ibuf->y - rymin;
if (xmax < 1 || ymax < 1) return;
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
/* The thing here is, the logic below (which was default behavior
* of how rectf is acquiring since forever) gives float buffer for
* composite output only. This buffer can not be used for other
* passes obviously.
*
* We might try finding corresponding for pass buffer in render result
* (which is actually missing when rendering with Cycles, who only
* writes all the passes when the tile is finished) or use float
* buffer from image buffer as reference, which is easier to use and
* contains all the data we need anyway.
* - sergey -
*/
/* TODO(sergey): Need to check has_combined here? */
if (iuser->pass == 0) {
RenderView *rv;
const int view_id = BKE_scene_multiview_view_id_get(&scene->r, viewname);
rv = RE_RenderViewGetById(rr, view_id);
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
/* find current float rect for display, first case is after composite... still weak */
if (rv->rectf)
rectf = rv->rectf;
else {
if (rv->rect32) {
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
/* special case, currently only happens with sequencer rendering,
* which updates the whole frame, so we can only mark display buffer
* as invalid here (sergey)
*/
ibuf->userflags |= IB_DISPLAY_BUFFER_INVALID;
return;
}
else {
Multi-View and Stereo 3D Official Documentation: http://www.blender.org/manual/render/workflows/multiview.html Implemented Features ==================== Builtin Stereo Camera * Convergence Mode * Interocular Distance * Convergence Distance * Pivot Mode Viewport * Cameras * Plane * Volume Compositor * View Switch Node * Image Node Multi-View OpenEXR support Sequencer * Image/Movie Strips 'Use Multiview' UV/Image Editor * Option to see Multi-View images in Stereo-3D or its individual images * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images I/O * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images Scene Render Views * Ability to have an arbitrary number of views in the scene Missing Bits ============ First rule of Multi-View bug report: If something is not working as it should *when Views is off* this is a severe bug, do mention this in the report. Second rule is, if something works *when Views is off* but doesn't (or crashes) when *Views is on*, this is a important bug. Do mention this in the report. Everything else is likely small todos, and may wait until we are sure none of the above is happening. Apart from that there are those known issues: * Compositor Image Node poorly working for Multi-View OpenEXR (this was working prefectly before the 'Use Multi-View' functionality) * Selecting camera from Multi-View when looking from camera is problematic * Animation Playback (ctrl+F11) doesn't support stereo formats * Wrong filepath when trying to play back animated scene * Viewport Rendering doesn't support Multi-View * Overscan Rendering * Fullscreen display modes need to warn the user * Object copy should be aware of views suffix Acknowledgments =============== * Francesco Siddi for the help with the original feature specs and design * Brecht Van Lommel for the original review of the code and design early on * Blender Foundation for the Development Fund to support the project wrap up Final patch reviewers: * Antony Riakiotakis (psy-fi) * Campbell Barton (ideasman42) * Julian Eisel (Severin) * Sergey Sharybin (nazgul) * Thomas Dinged (dingto) Code contributors of the original branch in github: * Alexey Akishin * Gabriel Caraballo
2015-04-06 10:40:12 -03:00
if (rr->renlay == NULL) return;
rectf = RE_RenderLayerGetPass(rr->renlay, RE_PASSNAME_COMBINED, viewname);
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
}
}
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
if (rectf == NULL) return;
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
rectf += 4 * (rr->rectx * ymin + xmin);
linear_stride = rr->rectx;
linear_offset_x = rxmin;
linear_offset_y = rymin;
}
else {
rectf = ibuf->rect_float;
linear_stride = ibuf->x;
linear_offset_x = 0;
linear_offset_y = 0;
}
if (rr->do_exr_tile) {
/* We don't support changing color management settings during rendering
* when using Save Buffers option.
*/
view_settings = &rj->view_settings;
display_settings = &rj->display_settings;
}
else {
view_settings = &scene->view_settings;
display_settings = &scene->display_settings;
}
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_partial_display_buffer_update(ibuf, rectf, NULL,
linear_stride, linear_offset_x, linear_offset_y,
view_settings, display_settings,
rxmin, rymin, rxmin + xmax, rymin + ymax,
rr->do_exr_tile);
}
/* ****************************** render invoking ***************** */
/* set callbacks, exported to sequence render too.
* Only call in foreground (UI) renders. */
static void screen_render_scene_layer_set(wmOperator *op, Main *mainp, Scene **scene, SceneRenderLayer **srl)
{
/* single layer re-render */
if (RNA_struct_property_is_set(op->ptr, "scene")) {
Scene *scn;
2012-03-29 22:42:32 +00:00
char scene_name[MAX_ID_NAME - 2];
RNA_string_get(op->ptr, "scene", scene_name);
scn = (Scene *)BLI_findstring(&mainp->scene, scene_name, offsetof(ID, name) + 2);
if (scn) {
/* camera switch wont have updated */
2012-03-29 22:42:32 +00:00
scn->r.cfra = (*scene)->r.cfra;
BKE_scene_camera_switch_update(scn);
*scene = scn;
}
}
if (RNA_struct_property_is_set(op->ptr, "layer")) {
SceneRenderLayer *rl;
char rl_name[RE_MAXNAME];
RNA_string_get(op->ptr, "layer", rl_name);
rl = (SceneRenderLayer *)BLI_findstring(&(*scene)->r.layers, rl_name, offsetof(SceneRenderLayer, name));
if (rl)
*srl = rl;
}
}
/* executes blocking render */
static int screen_render_exec(bContext *C, wmOperator *op)
{
2012-03-29 22:42:32 +00:00
Scene *scene = CTX_data_scene(C);
SceneRenderLayer *srl = NULL;
Render *re;
Image *ima;
2012-03-29 22:42:32 +00:00
View3D *v3d = CTX_wm_view3d(C);
Main *mainp = CTX_data_main(C);
unsigned int lay_override;
const bool is_animation = RNA_boolean_get(op->ptr, "animation");
const bool is_write_still = RNA_boolean_get(op->ptr, "write_still");
2012-03-29 22:42:32 +00:00
struct Object *camera_override = v3d ? V3D_CAMERA_LOCAL(v3d) : NULL;
/* custom scene and single layer re-render */
screen_render_scene_layer_set(op, mainp, &scene, &srl);
if (!is_animation && is_write_still && BKE_imtype_is_movie(scene->r.im_format.imtype)) {
BKE_report(op->reports, RPT_ERROR, "Cannot write a single file with an animation format selected");
return OPERATOR_CANCELLED;
}
re = RE_NewSceneRender(scene);
lay_override = (v3d && v3d->lay != scene->lay) ? v3d->lay : 0;
G.is_break = false;
RE_test_break_cb(re, NULL, render_break);
2012-03-29 22:42:32 +00:00
ima = BKE_image_verify_viewer(IMA_TYPE_R_RESULT, "Render Result");
BKE_image_signal(ima, NULL, IMA_SIGNAL_FREE);
BKE_image_backup_render(scene, ima, true);
/* cleanup sequencer caches before starting user triggered render.
2012-04-22 11:54:53 +00:00
* otherwise, invalidated cache entries can make their way into
* the output rendering. We can't put that into RE_BlenderFrame,
* since sequence rendering can call that recursively... (peter) */
BKE_sequencer_cache_cleanup();
RE_SetReports(re, op->reports);
BLI_threaded_malloc_begin();
if (is_animation)
RE_BlenderAnim(re, mainp, scene, camera_override, lay_override, scene->r.sfra, scene->r.efra, scene->r.frame_step);
else
RE_BlenderFrame(re, mainp, scene, srl, camera_override, lay_override, scene->r.cfra, is_write_still);
BLI_threaded_malloc_end();
RE_SetReports(re, NULL);
// no redraw needed, we leave state as we entered it
ED_update_for_newframe(mainp, scene, 1);
2012-03-29 22:42:32 +00:00
WM_event_add_notifier(C, NC_SCENE | ND_RENDER_RESULT, scene);
return OPERATOR_FINISHED;
}
static void render_freejob(void *rjv)
{
2012-03-29 22:42:32 +00:00
RenderJob *rj = rjv;
BKE_color_managed_view_settings_free(&rj->view_settings);
MEM_freeN(rj);
}
/* str is IMA_MAX_RENDER_TEXT in size */
static void make_renderinfo_string(const RenderStats *rs,
const Scene *scene,
const bool v3d_override,
const char *error,
char *str)
{
2012-03-29 22:42:32 +00:00
char info_time_str[32]; // used to be extern to header_info.c
uintptr_t mem_in_use, mmap_in_use, peak_memory;
float megs_used_memory, mmap_used_memory, megs_peak_memory;
2012-03-29 22:42:32 +00:00
char *spos = str;
2012-03-29 22:42:32 +00:00
mem_in_use = MEM_get_memory_in_use();
mmap_in_use = MEM_get_mapped_memory_in_use();
peak_memory = MEM_get_peak_memory();
2012-03-29 22:42:32 +00:00
megs_used_memory = (mem_in_use - mmap_in_use) / (1024.0 * 1024.0);
mmap_used_memory = (mmap_in_use) / (1024.0 * 1024.0);
megs_peak_memory = (peak_memory) / (1024.0 * 1024.0);
/* local view */
2013-06-25 10:44:30 +00:00
if (rs->localview)
spos += sprintf(spos, "%s | ", IFACE_("3D Local View"));
else if (v3d_override)
spos += sprintf(spos, "%s | ", IFACE_("3D View"));
/* frame number */
spos += sprintf(spos, IFACE_("Frame:%d "), (scene->r.cfra));
/* previous and elapsed time */
BLI_timecode_string_from_time_simple(info_time_str, sizeof(info_time_str), rs->lastframetime);
if (rs->infostr && rs->infostr[0]) {
if (rs->lastframetime != 0.0)
spos += sprintf(spos, IFACE_("| Last:%s "), info_time_str);
else
spos += sprintf(spos, "| ");
BLI_timecode_string_from_time_simple(info_time_str, sizeof(info_time_str), PIL_check_seconds_timer() - rs->starttime);
}
else
spos += sprintf(spos, "| ");
spos += sprintf(spos, IFACE_("Time:%s "), info_time_str);
/* statistics */
if (rs->statstr) {
2013-06-25 10:44:30 +00:00
if (rs->statstr[0]) {
spos += sprintf(spos, "| %s ", rs->statstr);
2013-06-25 10:44:30 +00:00
}
}
else {
if (rs->totvert || rs->totface || rs->tothalo || rs->totstrand || rs->totlamp)
spos += sprintf(spos, "| ");
if (rs->totvert) spos += sprintf(spos, IFACE_("Ve:%d "), rs->totvert);
if (rs->totface) spos += sprintf(spos, IFACE_("Fa:%d "), rs->totface);
if (rs->tothalo) spos += sprintf(spos, IFACE_("Ha:%d "), rs->tothalo);
if (rs->totstrand) spos += sprintf(spos, IFACE_("St:%d "), rs->totstrand);
if (rs->totlamp) spos += sprintf(spos, IFACE_("La:%d "), rs->totlamp);
if (rs->mem_peak == 0.0f)
spos += sprintf(spos, IFACE_("| Mem:%.2fM (%.2fM, Peak %.2fM) "),
megs_used_memory, mmap_used_memory, megs_peak_memory);
else
spos += sprintf(spos, IFACE_("| Mem:%.2fM, Peak: %.2fM "), rs->mem_used, rs->mem_peak);
if (rs->curfield)
spos += sprintf(spos, IFACE_("Field %d "), rs->curfield);
if (rs->curblur)
spos += sprintf(spos, IFACE_("Blur %d "), rs->curblur);
}
/* full sample */
if (rs->curfsa)
spos += sprintf(spos, IFACE_("| Full Sample %d "), rs->curfsa);
/* extra info */
if (rs->infostr && rs->infostr[0]) {
2012-03-29 22:42:32 +00:00
spos += sprintf(spos, "| %s ", rs->infostr);
}
else if (error && error[0]) {
spos += sprintf(spos, "| %s ", error);
}
/* very weak... but 512 characters is quite safe */
2012-03-29 22:42:32 +00:00
if (spos >= str + IMA_MAX_RENDER_TEXT)
if (G.debug & G_DEBUG)
printf("WARNING! renderwin text beyond limit\n");
}
static void image_renderinfo_cb(void *rjv, RenderStats *rs)
{
2012-03-29 22:42:32 +00:00
RenderJob *rj = rjv;
RenderResult *rr;
2012-03-29 22:42:32 +00:00
rr = RE_AcquireResultRead(rj->re);
if (rr) {
2010-07-05 09:31:24 +00:00
/* malloc OK here, stats_draw is not in tile threads */
2012-03-29 22:42:32 +00:00
if (rr->text == NULL)
rr->text = MEM_callocN(IMA_MAX_RENDER_TEXT, "rendertext");
2010-07-05 09:31:24 +00:00
make_renderinfo_string(rs, rj->scene, rj->v3d_override,
rr->error, rr->text);
2010-07-05 09:31:24 +00:00
}
RE_ReleaseResult(rj->re);
/* make jobs timer to send notifier */
*(rj->do_update) = true;
}
static void render_progress_update(void *rjv, float progress)
{
2012-03-29 22:42:32 +00:00
RenderJob *rj = rjv;
if (rj->progress && *rj->progress != progress) {
*rj->progress = progress;
/* make jobs timer to send notifier */
*(rj->do_update) = true;
}
}
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
/* Not totally reliable, but works fine in most of cases and
* in worst case would just make it so extra color management
* for the whole render result is applied (which was already
* happening already).
*/
static void render_image_update_pass_and_layer(RenderJob *rj, RenderResult *rr, ImageUser *iuser)
{
wmWindowManager *wm;
ScrArea *first_sa = NULL, *matched_sa = NULL;
/* image window, compo node users */
for (wm = rj->main->wm.first; wm && matched_sa == NULL; wm = wm->id.next) { /* only 1 wm */
wmWindow *win;
for (win = wm->windows.first; win && matched_sa == NULL; win = win->next) {
ScrArea *sa;
for (sa = win->screen->areabase.first; sa; sa = sa->next) {
if (sa->spacetype == SPACE_IMAGE) {
SpaceImage *sima = sa->spacedata.first;
// sa->spacedata might be empty when toggling fullscreen mode.
if (sima != NULL && sima->image == rj->image) {
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
if (first_sa == NULL) {
first_sa = sa;
}
if (sa == rj->sa) {
matched_sa = sa;
break;
}
}
}
}
}
}
if (matched_sa == NULL) {
matched_sa = first_sa;
}
if (matched_sa) {
SpaceImage *sima = matched_sa->spacedata.first;
RenderResult *main_rr = RE_AcquireResultRead(rj->re);
/* TODO(sergey): is there faster way to get the layer index? */
if (rr->renlay) {
int layer = BLI_findstringindex(&main_rr->layers,
(char *)rr->renlay->name,
offsetof(RenderLayer, name));
sima->iuser.layer = layer;
rj->last_layer = layer;
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
}
iuser->pass = sima->iuser.pass;
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
iuser->layer = sima->iuser.layer;
RE_ReleaseResult(rj->re);
}
}
static void image_rect_update(void *rjv, RenderResult *rr, volatile rcti *renrect)
{
2012-03-29 22:42:32 +00:00
RenderJob *rj = rjv;
Image *ima = rj->image;
ImBuf *ibuf;
void *lock;
Multi-View and Stereo 3D Official Documentation: http://www.blender.org/manual/render/workflows/multiview.html Implemented Features ==================== Builtin Stereo Camera * Convergence Mode * Interocular Distance * Convergence Distance * Pivot Mode Viewport * Cameras * Plane * Volume Compositor * View Switch Node * Image Node Multi-View OpenEXR support Sequencer * Image/Movie Strips 'Use Multiview' UV/Image Editor * Option to see Multi-View images in Stereo-3D or its individual images * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images I/O * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images Scene Render Views * Ability to have an arbitrary number of views in the scene Missing Bits ============ First rule of Multi-View bug report: If something is not working as it should *when Views is off* this is a severe bug, do mention this in the report. Second rule is, if something works *when Views is off* but doesn't (or crashes) when *Views is on*, this is a important bug. Do mention this in the report. Everything else is likely small todos, and may wait until we are sure none of the above is happening. Apart from that there are those known issues: * Compositor Image Node poorly working for Multi-View OpenEXR (this was working prefectly before the 'Use Multi-View' functionality) * Selecting camera from Multi-View when looking from camera is problematic * Animation Playback (ctrl+F11) doesn't support stereo formats * Wrong filepath when trying to play back animated scene * Viewport Rendering doesn't support Multi-View * Overscan Rendering * Fullscreen display modes need to warn the user * Object copy should be aware of views suffix Acknowledgments =============== * Francesco Siddi for the help with the original feature specs and design * Brecht Van Lommel for the original review of the code and design early on * Blender Foundation for the Development Fund to support the project wrap up Final patch reviewers: * Antony Riakiotakis (psy-fi) * Campbell Barton (ideasman42) * Julian Eisel (Severin) * Sergey Sharybin (nazgul) * Thomas Dinged (dingto) Code contributors of the original branch in github: * Alexey Akishin * Gabriel Caraballo
2015-04-06 10:40:12 -03:00
const char *viewname = RE_GetActiveRenderView(rj->re);
/* only update if we are displaying the slot being rendered */
if (ima->render_slot != ima->last_render_slot) {
rj->image_outdated = true;
return;
}
else if (rj->image_outdated) {
/* update entire render */
rj->image_outdated = false;
BKE_image_signal(ima, NULL, IMA_SIGNAL_COLORMANAGE);
*(rj->do_update) = true;
return;
}
if (rr == NULL)
return;
/* update part of render */
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
render_image_update_pass_and_layer(rj, rr, &rj->iuser);
2012-03-29 22:42:32 +00:00
ibuf = BKE_image_acquire_ibuf(ima, &rj->iuser, &lock);
if (ibuf) {
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
/* Don't waste time on CPU side color management if
* image will be displayed using GLSL.
*
* Need to update rect if Save Buffers enabled because in
* this case GLSL doesn't have original float buffer to
* operate with.
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
*/
if (rr->do_exr_tile ||
!rj->supports_glsl_draw ||
ibuf->channels == 1 ||
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
U.image_draw_method != IMAGE_DRAW_METHOD_GLSL)
{
Multi-View and Stereo 3D Official Documentation: http://www.blender.org/manual/render/workflows/multiview.html Implemented Features ==================== Builtin Stereo Camera * Convergence Mode * Interocular Distance * Convergence Distance * Pivot Mode Viewport * Cameras * Plane * Volume Compositor * View Switch Node * Image Node Multi-View OpenEXR support Sequencer * Image/Movie Strips 'Use Multiview' UV/Image Editor * Option to see Multi-View images in Stereo-3D or its individual images * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images I/O * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images Scene Render Views * Ability to have an arbitrary number of views in the scene Missing Bits ============ First rule of Multi-View bug report: If something is not working as it should *when Views is off* this is a severe bug, do mention this in the report. Second rule is, if something works *when Views is off* but doesn't (or crashes) when *Views is on*, this is a important bug. Do mention this in the report. Everything else is likely small todos, and may wait until we are sure none of the above is happening. Apart from that there are those known issues: * Compositor Image Node poorly working for Multi-View OpenEXR (this was working prefectly before the 'Use Multi-View' functionality) * Selecting camera from Multi-View when looking from camera is problematic * Animation Playback (ctrl+F11) doesn't support stereo formats * Wrong filepath when trying to play back animated scene * Viewport Rendering doesn't support Multi-View * Overscan Rendering * Fullscreen display modes need to warn the user * Object copy should be aware of views suffix Acknowledgments =============== * Francesco Siddi for the help with the original feature specs and design * Brecht Van Lommel for the original review of the code and design early on * Blender Foundation for the Development Fund to support the project wrap up Final patch reviewers: * Antony Riakiotakis (psy-fi) * Campbell Barton (ideasman42) * Julian Eisel (Severin) * Sergey Sharybin (nazgul) * Thomas Dinged (dingto) Code contributors of the original branch in github: * Alexey Akishin * Gabriel Caraballo
2015-04-06 10:40:12 -03:00
image_buffer_rect_update(rj, rr, ibuf, &rj->iuser, renrect, viewname);
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
}
/* make jobs timer to send notifier */
*(rj->do_update) = true;
}
BKE_image_release_ibuf(ima, ibuf, lock);
}
static void current_scene_update(void *rjv, Scene *scene)
{
RenderJob *rj = rjv;
rj->current_scene = scene;
rj->iuser.scene = scene;
}
static void render_startjob(void *rjv, short *stop, short *do_update, float *progress)
{
2012-03-29 22:42:32 +00:00
RenderJob *rj = rjv;
2012-03-29 22:42:32 +00:00
rj->stop = stop;
rj->do_update = do_update;
rj->progress = progress;
RE_SetReports(rj->re, rj->reports);
if (rj->anim)
RE_BlenderAnim(rj->re, rj->main, rj->scene, rj->camera_override, rj->lay_override, rj->scene->r.sfra, rj->scene->r.efra, rj->scene->r.frame_step);
else
RE_BlenderFrame(rj->re, rj->main, rj->scene, rj->srl, rj->camera_override, rj->lay_override, rj->scene->r.cfra, rj->write_still);
RE_SetReports(rj->re, NULL);
}
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
static void render_image_restore_layer(RenderJob *rj)
{
wmWindowManager *wm;
/* image window, compo node users */
for (wm = rj->main->wm.first; wm; wm = wm->id.next) { /* only 1 wm */
wmWindow *win;
for (win = wm->windows.first; win; win = win->next) {
ScrArea *sa;
for (sa = win->screen->areabase.first; sa; sa = sa->next) {
if (sa == rj->sa) {
if (sa->spacetype == SPACE_IMAGE) {
SpaceImage *sima = sa->spacedata.first;
if (RE_HasSingleLayer(rj->re)) {
/* For single layer renders keep the active layer
* visible, or show the compositing result. */
RenderResult *rr = RE_AcquireResultRead(rj->re);
if(RE_HasCombinedLayer(rr)) {
sima->iuser.layer = 0;
}
RE_ReleaseResult(rj->re);
}
else {
/* For multiple layer render, set back the layer
* that was set at the start of rendering. */
sima->iuser.layer = rj->orig_layer;
}
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
}
return;
}
}
}
}
}
static void render_endjob(void *rjv)
{
2012-03-29 22:42:32 +00:00
RenderJob *rj = rjv;
/* this render may be used again by the sequencer without the active 'Render' where the callbacks
* would be re-assigned. assign dummy callbacks to avoid referencing freed renderjobs bug [#24508] */
RE_InitRenderCB(rj->re);
if (rj->main != G.main)
BKE_main_free(rj->main);
/* else the frame will not update for the original value */
if (rj->anim && !(rj->scene->r.scemode & R_NO_FRAME_UPDATE)) {
/* possible this fails of loading new file while rendering */
if (G.main->wm.first) {
ED_update_for_newframe(G.main, rj->scene, 1);
}
}
/* XXX above function sets all tags in nodes */
ntreeCompositClearTags(rj->scene->nodetree);
/* potentially set by caller */
rj->scene->r.scemode &= ~R_NO_FRAME_UPDATE;
if (rj->srl) {
nodeUpdateID(rj->scene->nodetree, &rj->scene->id);
2012-03-29 22:42:32 +00:00
WM_main_add_notifier(NC_NODE | NA_EDITED, rj->scene);
}
Bunch of fixes for GLSL display transform - GLSL shader wasn't aware of alpha predivide option, always assuming alpha is straight. Gave wrong results when displaying transparent float buffers. - GLSL display wasn't aware of float buffers with number of channels different from 4, crashing when trying to display image with different number of channels. This required a bit larger changes, namely now it's possible to pass format (GL_RGB, GL_RGBAm GL_LUMINANCE) to glaDrawPixelsTex, This also implied adding format to glaDrawPixelsAuto and modifying all places where this functions are called. Now GLSL will handle both 3 and 4 channels buffers, single channel images are handled by CPU. - Replaced hack for render result displaying with a bit different hack. Namely CPU conversion will happen only during render, once render is done GLSL would be used for displaying render result on a screen. This is so because of the way renderer updates parts of the image -- it happens without respect to active render layer in image user. This is harmless because only display buffer is modifying, but this is tricky because we don't have original buffer opened during rendering. One more related fix here was about when rendering multiple layers, wrong image would be displaying when rendering is done. Added a signal to invalidate display buffer once rendering is done (only happens when using multiple layers). This solves issue with wrong buffer stuck on the display when using regular CPU display space transform and if GLSL is available it'll make image displayed with a GLSL shader. - As an additional change, byte buffers now also uses GLSL display transform. So now only dutehr and RGB curves are stoppers for using GLSL for all kind of display transforms.
2013-04-03 15:59:54 +00:00
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
if (rj->sa) {
render_image_restore_layer(rj);
}
/* XXX render stability hack */
G.is_rendering = false;
2013-06-25 10:44:30 +00:00
WM_main_add_notifier(NC_SCENE | ND_RENDER_RESULT, NULL);
Bunch of fixes for GLSL display transform - GLSL shader wasn't aware of alpha predivide option, always assuming alpha is straight. Gave wrong results when displaying transparent float buffers. - GLSL display wasn't aware of float buffers with number of channels different from 4, crashing when trying to display image with different number of channels. This required a bit larger changes, namely now it's possible to pass format (GL_RGB, GL_RGBAm GL_LUMINANCE) to glaDrawPixelsTex, This also implied adding format to glaDrawPixelsAuto and modifying all places where this functions are called. Now GLSL will handle both 3 and 4 channels buffers, single channel images are handled by CPU. - Replaced hack for render result displaying with a bit different hack. Namely CPU conversion will happen only during render, once render is done GLSL would be used for displaying render result on a screen. This is so because of the way renderer updates parts of the image -- it happens without respect to active render layer in image user. This is harmless because only display buffer is modifying, but this is tricky because we don't have original buffer opened during rendering. One more related fix here was about when rendering multiple layers, wrong image would be displaying when rendering is done. Added a signal to invalidate display buffer once rendering is done (only happens when using multiple layers). This solves issue with wrong buffer stuck on the display when using regular CPU display space transform and if GLSL is available it'll make image displayed with a GLSL shader. - As an additional change, byte buffers now also uses GLSL display transform. So now only dutehr and RGB curves are stoppers for using GLSL for all kind of display transforms.
2013-04-03 15:59:54 +00:00
/* Partial render result will always update display buffer
* for first render layer only. This is nice because you'll
* see render progress during rendering, but it ends up in
* wrong display buffer shown after rendering.
*
* The code below will mark display buffer as invalid after
* rendering in case multiple layers were rendered, which
* ensures display buffer matches render layer after
* rendering.
*
* Perhaps proper way would be to toggle active render
* layer in image editor and job, so we always display
* layer being currently rendered. But this is not so much
* trivial at this moment, especially because of external
* engine API, so lets use simple and robust way for now
* - sergey -
*/
if (rj->scene->r.layers.first != rj->scene->r.layers.last ||
rj->image_outdated)
{
Bunch of fixes for GLSL display transform - GLSL shader wasn't aware of alpha predivide option, always assuming alpha is straight. Gave wrong results when displaying transparent float buffers. - GLSL display wasn't aware of float buffers with number of channels different from 4, crashing when trying to display image with different number of channels. This required a bit larger changes, namely now it's possible to pass format (GL_RGB, GL_RGBAm GL_LUMINANCE) to glaDrawPixelsTex, This also implied adding format to glaDrawPixelsAuto and modifying all places where this functions are called. Now GLSL will handle both 3 and 4 channels buffers, single channel images are handled by CPU. - Replaced hack for render result displaying with a bit different hack. Namely CPU conversion will happen only during render, once render is done GLSL would be used for displaying render result on a screen. This is so because of the way renderer updates parts of the image -- it happens without respect to active render layer in image user. This is harmless because only display buffer is modifying, but this is tricky because we don't have original buffer opened during rendering. One more related fix here was about when rendering multiple layers, wrong image would be displaying when rendering is done. Added a signal to invalidate display buffer once rendering is done (only happens when using multiple layers). This solves issue with wrong buffer stuck on the display when using regular CPU display space transform and if GLSL is available it'll make image displayed with a GLSL shader. - As an additional change, byte buffers now also uses GLSL display transform. So now only dutehr and RGB curves are stoppers for using GLSL for all kind of display transforms.
2013-04-03 15:59:54 +00:00
void *lock;
Image *ima = rj->image;
ImBuf *ibuf = BKE_image_acquire_ibuf(ima, &rj->iuser, &lock);
if (ibuf)
ibuf->userflags |= IB_DISPLAY_BUFFER_INVALID;
BKE_image_release_ibuf(ima, ibuf, lock);
}
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
/* Finally unlock the user interface (if it was locked). */
if (rj->interface_locked) {
Scene *scene;
/* Interface was locked, so window manager couldn't have been changed
* and using one from Global will unlock exactly the same manager as
* was locked before running the job.
*/
WM_set_locked_interface(G.main->wm.first, false);
/* We've freed all the derived caches before rendering, which is
* effectively the same as if we re-loaded the file.
*
* So let's not try being smart here and just reset all updated
* scene layers and use generic DAG_on_visible_update.
*/
for (scene = G.main->scene.first; scene; scene = scene->id.next) {
scene->lay_updated = 0;
}
DAG_on_visible_update(G.main, false);
}
}
/* called by render, check job 'stop' value or the global */
static int render_breakjob(void *rjv)
{
2012-03-29 22:42:32 +00:00
RenderJob *rj = rjv;
if (G.is_break)
return 1;
if (rj->stop && *(rj->stop))
return 1;
return 0;
}
/* for exec() when there is no render job
* note: this wont check for the escape key being pressed, but doing so isnt threadsafe */
static int render_break(void *UNUSED(rjv))
{
if (G.is_break)
return 1;
return 0;
}
/* runs in thread, no cursor setting here works. careful with notifiers too (malloc conflicts) */
/* maybe need a way to get job send notifer? */
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
static void render_drawlock(void *rjv, int lock)
{
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
RenderJob *rj = rjv;
/* If interface is locked, renderer callback shall do nothing. */
if (!rj->interface_locked) {
BKE_spacedata_draw_locks(lock);
}
}
/* catch esc */
static int screen_render_modal(bContext *C, wmOperator *op, const wmEvent *event)
{
Scene *scene = (Scene *) op->customdata;
/* no running blender, remove handler and pass through */
if (0 == WM_jobs_test(CTX_wm_manager(C), scene, WM_JOB_TYPE_RENDER)) {
2012-03-29 22:42:32 +00:00
return OPERATOR_FINISHED | OPERATOR_PASS_THROUGH;
}
/* running render */
switch (event->type) {
case ESCKEY:
return OPERATOR_RUNNING_MODAL;
}
return OPERATOR_PASS_THROUGH;
}
static void screen_render_cancel(bContext *C, wmOperator *op)
{
wmWindowManager *wm = CTX_wm_manager(C);
Scene *scene = (Scene *) op->customdata;
/* kill on cancel, because job is using op->reports */
WM_jobs_kill_type(wm, scene, WM_JOB_TYPE_RENDER);
}
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
static void clean_viewport_memory(Main *bmain, Scene *scene, int renderlay)
{
Object *object;
Scene *sce_iter;
Base *base;
for (object = bmain->object.first; object; object = object->id.next) {
object->id.tag |= LIB_TAG_DOIT;
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
}
for (SETLOOPER(scene, sce_iter, base)) {
if ((base->lay & renderlay) == 0) {
continue;
}
if (RE_allow_render_generic_object(base->object)) {
base->object->id.tag &= ~LIB_TAG_DOIT;
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
}
}
for (SETLOOPER(scene, sce_iter, base)) {
object = base->object;
if ((object->id.tag & LIB_TAG_DOIT) == 0) {
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
continue;
}
object->id.tag &= ~LIB_TAG_DOIT;
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
BKE_object_free_derived_caches(object);
}
}
/* using context, starts job */
static int screen_render_invoke(bContext *C, wmOperator *op, const wmEvent *event)
{
/* new render clears all callbacks */
Main *mainp;
2012-03-29 22:42:32 +00:00
Scene *scene = CTX_data_scene(C);
SceneRenderLayer *srl = NULL;
Render *re;
wmJob *wm_job;
RenderJob *rj;
Image *ima;
int jobflag;
const bool is_animation = RNA_boolean_get(op->ptr, "animation");
const bool is_write_still = RNA_boolean_get(op->ptr, "write_still");
const bool use_viewport = RNA_boolean_get(op->ptr, "use_viewport");
View3D *v3d = use_viewport ? CTX_wm_view3d(C) : NULL;
2012-03-29 22:42:32 +00:00
struct Object *camera_override = v3d ? V3D_CAMERA_LOCAL(v3d) : NULL;
const char *name;
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
ScrArea *sa;
/* only one render job at a time */
if (WM_jobs_test(CTX_wm_manager(C), scene, WM_JOB_TYPE_RENDER))
return OPERATOR_CANCELLED;
if (RE_force_single_renderlayer(scene))
WM_event_add_notifier(C, NC_SCENE | ND_RENDER_OPTIONS, NULL);
if (!RE_is_rendering_allowed(scene, camera_override, op->reports)) {
return OPERATOR_CANCELLED;
}
if (!is_animation && is_write_still && BKE_imtype_is_movie(scene->r.im_format.imtype)) {
BKE_report(op->reports, RPT_ERROR, "Cannot write a single file with an animation format selected");
return OPERATOR_CANCELLED;
}
/* stop all running jobs, except screen one. currently previews frustrate Render */
WM_jobs_kill_all_except(CTX_wm_manager(C), CTX_wm_screen(C));
/* get main */
if (G.debug_value == 101) {
/* thread-safety experiment, copy main from the undo buffer */
struct MemFile *memfile = ED_undosys_stack_memfile_get_active(CTX_wm_manager(C)->undo_stack);
mainp = BLO_memfile_main_get(memfile, CTX_data_main(C), &scene);
}
else
2012-03-29 22:42:32 +00:00
mainp = CTX_data_main(C);
/* cancel animation playback */
if (ED_screen_animation_playing(CTX_wm_manager(C)))
ED_screen_animation_play(C, 0, 0);
/* handle UI stuff */
WM_cursor_wait(1);
/* flush sculpt and editmode changes */
ED_editors_flush_edits(C, true);
/* cleanup sequencer caches before starting user triggered render.
2012-04-22 11:54:53 +00:00
* otherwise, invalidated cache entries can make their way into
* the output rendering. We can't put that into RE_BlenderFrame,
* since sequence rendering can call that recursively... (peter) */
BKE_sequencer_cache_cleanup();
// store spare
// get view3d layer, local layer, make this nice api call to render
// store spare
/* ensure at least 1 area shows result */
sa = render_view_open(C, event->x, event->y, op->reports);
2012-03-29 22:42:32 +00:00
jobflag = WM_JOB_EXCL_RENDER | WM_JOB_PRIORITY | WM_JOB_PROGRESS;
/* custom scene and single layer re-render */
screen_render_scene_layer_set(op, mainp, &scene, &srl);
if (RNA_struct_property_is_set(op->ptr, "layer"))
jobflag |= WM_JOB_SUSPEND;
/* job custom data */
2012-03-29 22:42:32 +00:00
rj = MEM_callocN(sizeof(RenderJob), "render job");
rj->main = mainp;
rj->scene = scene;
rj->current_scene = rj->scene;
rj->srl = srl;
rj->camera_override = camera_override;
rj->lay_override = 0;
2012-03-29 22:42:32 +00:00
rj->anim = is_animation;
rj->write_still = is_write_still && !is_animation;
rj->iuser.scene = scene;
rj->iuser.ok = 1;
rj->reports = op->reports;
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
rj->orig_layer = 0;
rj->last_layer = 0;
rj->sa = sa;
rj->supports_glsl_draw = IMB_colormanagement_support_glsl_draw(&scene->view_settings);
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
BKE_color_managed_display_settings_copy(&rj->display_settings, &scene->display_settings);
BKE_color_managed_view_settings_copy(&rj->view_settings, &scene->view_settings);
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
if (sa) {
SpaceImage *sima = sa->spacedata.first;
rj->orig_layer = sima->iuser.layer;
}
2012-12-01 07:58:27 +00:00
if (v3d) {
if (scene->lay != v3d->lay) {
rj->lay_override = v3d->lay;
rj->v3d_override = true;
}
else if (camera_override && camera_override != scene->camera)
rj->v3d_override = true;
2012-12-01 07:58:27 +00:00
if (v3d->localvd)
rj->lay_override |= v3d->localvd->lay;
}
Option to lock the interface while rendering Added function called WM_set_locked_interface which does two things: - Prevents event queue from being handled, so no operators (see below) or values are even possible to run or change. This prevents any kind of "destructive" action performed from user while rendering. - Locks interface refresh for regions which does have lock set to truth in their template. Currently it's just a 3D viewport, but in the future more regions could be considered unsafe, or we could want to lock different parts of interface when doing different jobs. This is needed because 3D viewport could be using or changing the same data as renderer currently uses, leading to threading conflict. Notifiers are still allowed to handle, so render progress is seen on the screen, but would need to doublecheck on this, in terms some notifiers could be changing the data. For now interface locking happens for render job only in case "Lock Interface" checkbox is enabled. Other tools like backing would also benefit of this option. It is possible to mark operator as safe to be used in locked interface mode by adding OPTYPE_ALLOW_LOCKED bit to operator template flags. This bit is completely handled by wm_evem_system, not with operator run routines, so it's still possible to run operators from drivers and handlers. Currently allowed image editor navigation and zooming. Reviewers: brecht, campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D142
2014-01-29 16:07:14 +06:00
/* Lock the user interface depending on render settings. */
if (scene->r.use_lock_interface) {
int renderlay = rj->lay_override ? rj->lay_override : scene->lay;
WM_set_locked_interface(CTX_wm_manager(C), true);
/* Set flag interface need to be unlocked.
*
* This is so because we don't have copy of render settings
* accessible from render job and copy is needed in case
* of non-locked rendering, so we wouldn't try to unlock
* anything if option was initially unset but then was
* enabled during rendering.
*/
rj->interface_locked = true;
/* Clean memory used by viewport? */
clean_viewport_memory(rj->main, scene, renderlay);
}
/* setup job */
2012-03-29 22:42:32 +00:00
if (RE_seq_render_active(scene, &scene->r)) name = "Sequence Render";
else name = "Render";
wm_job = WM_jobs_get(CTX_wm_manager(C), CTX_wm_window(C), scene, name, jobflag, WM_JOB_TYPE_RENDER);
WM_jobs_customdata_set(wm_job, rj, render_freejob);
WM_jobs_timer(wm_job, 0.2, NC_SCENE | ND_RENDER_RESULT, 0);
WM_jobs_callbacks(wm_job, render_startjob, NULL, NULL, render_endjob);
/* get a render result image, and make sure it is empty */
2012-03-29 22:42:32 +00:00
ima = BKE_image_verify_viewer(IMA_TYPE_R_RESULT, "Render Result");
BKE_image_signal(ima, NULL, IMA_SIGNAL_FREE);
BKE_image_backup_render(rj->scene, ima, true);
2012-03-29 22:42:32 +00:00
rj->image = ima;
/* setup new render */
re = RE_NewSceneRender(scene);
RE_test_break_cb(re, rj, render_breakjob);
RE_draw_lock_cb(re, rj, render_drawlock);
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
RE_display_update_cb(re, rj, image_rect_update);
RE_current_scene_update_cb(re, rj, current_scene_update);
RE_stats_draw_cb(re, rj, image_renderinfo_cb);
RE_progress_cb(re, rj, render_progress_update);
2012-03-29 22:42:32 +00:00
rj->re = re;
G.is_break = false;
/* store actual owner of job, so modal operator could check for it,
* the reason of this is that active scene could change when rendering
2012-08-24 23:22:34 +00:00
* several layers from compositor [#31800]
*/
op->customdata = scene;
WM_jobs_start(CTX_wm_manager(C), wm_job);
WM_cursor_wait(0);
2012-03-29 22:42:32 +00:00
WM_event_add_notifier(C, NC_SCENE | ND_RENDER_RESULT, scene);
/* we set G.is_rendering here already instead of only in the job, this ensure
* main loop or other scene updates are disabled in time, since they may
* have started before the job thread */
G.is_rendering = true;
/* add modal handler for ESC */
WM_event_add_modal_handler(C, op);
return OPERATOR_RUNNING_MODAL;
}
/* contextual render, using current scene, view3d? */
void RENDER_OT_render(wmOperatorType *ot)
{
PropertyRNA *prop;
/* identifiers */
ot->name = "Render";
ot->description = "Render active scene";
ot->idname = "RENDER_OT_render";
/* api callbacks */
ot->invoke = screen_render_invoke;
ot->modal = screen_render_modal;
ot->cancel = screen_render_cancel;
ot->exec = screen_render_exec;
/*ot->poll = ED_operator_screenactive;*/ /* this isn't needed, causes failer in background mode */
prop = RNA_def_boolean(ot->srna, "animation", 0, "Animation", "Render files from the animation range of this scene");
RNA_def_property_flag(prop, PROP_SKIP_SAVE);
RNA_def_boolean(ot->srna, "write_still", 0, "Write Image", "Save rendered the image to the output path (used only when animation is disabled)");
prop = RNA_def_boolean(ot->srna, "use_viewport", 0, "Use 3D Viewport", "When inside a 3D viewport, use layers and camera of the viewport");
RNA_def_property_flag(prop, PROP_SKIP_SAVE);
prop = RNA_def_string(ot->srna, "layer", NULL, RE_MAXNAME, "Render Layer", "Single render layer to re-render (used only when animation is disabled)");
RNA_def_property_flag(prop, PROP_SKIP_SAVE);
prop = RNA_def_string(ot->srna, "scene", NULL, MAX_ID_NAME - 2, "Scene", "Scene to render, current scene if not specified");
RNA_def_property_flag(prop, PROP_SKIP_SAVE);
}
/* ************** preview for 3d viewport ***************** */
#define PR_UPDATE_VIEW 1
#define PR_UPDATE_RENDERSIZE 2
#define PR_UPDATE_MATERIAL 4
#define PR_UPDATE_DATABASE 8
typedef struct RenderPreview {
/* from wmJob */
void *owner;
short *stop, *do_update;
wmJob *job;
Scene *scene;
ScrArea *sa;
ARegion *ar;
View3D *v3d;
RegionView3D *rv3d;
Main *bmain;
RenderEngine *engine;
float viewmat[4][4];
int start_resolution_divider;
int resolution_divider;
bool has_freestyle;
} RenderPreview;
static int render_view3d_disprect(Scene *scene, ARegion *ar, View3D *v3d, RegionView3D *rv3d, rcti *disprect)
{
/* copied code from view3d_draw.c */
rctf viewborder;
int draw_border;
if (rv3d->persp == RV3D_CAMOB)
draw_border = (scene->r.mode & R_BORDER) != 0;
else
draw_border = (v3d->flag2 & V3D_RENDER_BORDER) != 0;
if (draw_border) {
if (rv3d->persp == RV3D_CAMOB) {
ED_view3d_calc_camera_border(scene, ar, v3d, rv3d, &viewborder, false);
disprect->xmin = viewborder.xmin + scene->r.border.xmin * BLI_rctf_size_x(&viewborder);
disprect->ymin = viewborder.ymin + scene->r.border.ymin * BLI_rctf_size_y(&viewborder);
disprect->xmax = viewborder.xmin + scene->r.border.xmax * BLI_rctf_size_x(&viewborder);
disprect->ymax = viewborder.ymin + scene->r.border.ymax * BLI_rctf_size_y(&viewborder);
}
else {
disprect->xmin = v3d->render_border.xmin * ar->winx;
disprect->xmax = v3d->render_border.xmax * ar->winx;
disprect->ymin = v3d->render_border.ymin * ar->winy;
disprect->ymax = v3d->render_border.ymax * ar->winy;
}
return 1;
}
BLI_rcti_init(disprect, 0, 0, 0, 0);
return 0;
}
2013-04-17 01:16:23 +00:00
/* returns true if OK */
static bool render_view3d_get_rects(ARegion *ar, View3D *v3d, RegionView3D *rv3d, rctf *viewplane, RenderEngine *engine,
float *r_clipsta, float *r_clipend, float *r_pixsize, bool *r_ortho)
{
2013-04-17 01:16:23 +00:00
if (ar->winx < 4 || ar->winy < 4) return false;
*r_ortho = ED_view3d_viewplane_get(v3d, rv3d, ar->winx, ar->winy, viewplane, r_clipsta, r_clipend, r_pixsize);
engine->resolution_x = ar->winx;
engine->resolution_y = ar->winy;
2013-04-17 01:16:23 +00:00
return true;
}
static bool render_view3d_is_valid(RenderPreview *rp)
{
return (rp->rv3d->render_engine != NULL);
}
/* called by renderer, checks job value */
static int render_view3d_break(void *rpv)
{
RenderPreview *rp = rpv;
if (G.is_break)
return 1;
/* during render, rv3d->engine can get freed */
if (render_view3d_is_valid(rp) == false) {
*rp->stop = 1;
}
return *(rp->stop);
}
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
static void render_view3d_display_update(void *rpv, RenderResult *UNUSED(rr), volatile struct rcti *UNUSED(rect))
{
RenderPreview *rp = rpv;
*(rp->do_update) = true;
}
static void render_view3d_renderinfo_cb(void *rjp, RenderStats *rs)
{
RenderPreview *rp = rjp;
/* during render, rv3d->engine can get freed */
if (rp->rv3d->render_engine == NULL) {
*rp->stop = 1;
}
else {
make_renderinfo_string(rs, rp->scene, false, NULL, rp->engine->text);
/* make jobs timer to send notifier */
*(rp->do_update) = true;
}
}
BLI_INLINE void rcti_scale_coords(rcti *scaled_rect, const rcti *rect,
const float scale)
{
scaled_rect->xmin = rect->xmin * scale;
scaled_rect->ymin = rect->ymin * scale;
scaled_rect->xmax = rect->xmax * scale;
scaled_rect->ymax = rect->ymax * scale;
}
static void render_update_resolution(Render *re, const RenderPreview *rp,
bool use_border, const rcti *clip_rect)
{
2015-02-11 18:38:41 +11:00
int winx = rp->ar->winx / rp->resolution_divider;
int winy = rp->ar->winy / rp->resolution_divider;
if (use_border) {
rcti scaled_cliprct;
rcti_scale_coords(&scaled_cliprct, clip_rect,
1.0f / rp->resolution_divider);
RE_ChangeResolution(re, winx, winy, &scaled_cliprct);
}
else {
RE_ChangeResolution(re, winx, winy, NULL);
}
if (rp->has_freestyle) {
if (rp->resolution_divider == BKE_render_preview_pixel_size(&rp->scene->r)) {
RE_ChangeModeFlag(re, R_EDGE_FRS, false);
}
else {
RE_ChangeModeFlag(re, R_EDGE_FRS, true);
}
}
}
static void render_view3d_startjob(void *customdata, short *stop, short *do_update, float *UNUSED(progress))
{
RenderPreview *rp = customdata;
Render *re;
RenderStats *rstats;
rctf viewplane;
rcti cliprct;
float clipsta, clipend, pixsize;
2013-04-17 01:16:23 +00:00
bool orth, restore = 0;
char name[32];
int update_flag;
bool use_border;
int ob_inst_update_flag = 0;
update_flag = rp->engine->job_update_flag;
rp->engine->job_update_flag = 0;
//printf("ma %d res %d view %d db %d\n", update_flag & PR_UPDATE_MATERIAL, update_flag & PR_UPDATE_RENDERSIZE, update_flag & PR_UPDATE_VIEW, update_flag & PR_UPDATE_DATABASE);
G.is_break = false;
if (false == render_view3d_get_rects(rp->ar, rp->v3d, rp->rv3d, &viewplane, rp->engine, &clipsta, &clipend, &pixsize, &orth))
return;
rp->stop = stop;
rp->do_update = do_update;
// printf("Enter previewrender\n");
/* ok, are we rendering all over? */
sprintf(name, "View3dPreview %p", (void *)rp->ar);
2013-04-17 01:16:23 +00:00
re = rp->engine->re = RE_GetRender(name);
/* set this always, rp is different for each job */
RE_test_break_cb(re, rp, render_view3d_break);
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
RE_display_update_cb(re, rp, render_view3d_display_update);
RE_stats_draw_cb(re, rp, render_view3d_renderinfo_cb);
2013-04-17 01:16:23 +00:00
rstats = RE_GetStats(re);
if (update_flag & PR_UPDATE_VIEW) {
Object *object;
rp->resolution_divider = rp->start_resolution_divider;
/* Same as database_init_objects(), loop over all objects.
* We might consider de-duplicating the code between this two cases.
*/
for (object = rp->bmain->object.first; object; object = object->id.next) {
float mat[4][4];
mul_m4_m4m4(mat, rp->viewmat, object->obmat);
invert_m4_m4(object->imat_ren, mat);
2014-08-17 12:18:40 +10:00
}
}
use_border = render_view3d_disprect(rp->scene, rp->ar, rp->v3d,
rp->rv3d, &cliprct);
if ((update_flag & (PR_UPDATE_RENDERSIZE | PR_UPDATE_DATABASE | PR_UPDATE_VIEW)) || rstats->convertdone == 0) {
RenderData rdata;
/* no osa, blur, seq, layers, savebuffer etc for preview render */
rdata = rp->scene->r;
rdata.mode &= ~(R_OSA | R_MBLUR | R_BORDER | R_PANORAMA);
rdata.scemode &= ~(R_DOSEQ | R_DOCOMP | R_FREE_IMAGE | R_EXR_TILE_FILE | R_FULL_SAMPLE);
rdata.scemode |= R_VIEWPORT_PREVIEW;
/* we do use layers, but only active */
rdata.scemode |= R_SINGLE_LAYER;
/* initalize always */
if (use_border) {
rdata.mode |= R_BORDER;
RE_InitState(re, NULL, &rdata, NULL, rp->ar->winx, rp->ar->winy, &cliprct);
}
else
RE_InitState(re, NULL, &rdata, NULL, rp->ar->winx, rp->ar->winy, NULL);
}
if (orth)
RE_SetOrtho(re, &viewplane, clipsta, clipend);
else
RE_SetWindow(re, &viewplane, clipsta, clipend);
RE_SetPixelSize(re, pixsize);
if ((update_flag & PR_UPDATE_DATABASE) || rstats->convertdone == 0) {
unsigned int lay = rp->scene->lay;
/* allow localview render for objects with lights in normal layers */
if (rp->v3d->lay & 0xFF000000)
lay |= rp->v3d->lay;
else lay = rp->v3d->lay;
RE_SetView(re, rp->viewmat);
/* copying blender data while main thread is locked, to avoid crashes */
WM_job_main_thread_lock_acquire(rp->job);
RE_Database_Free(re);
RE_Database_FromScene(re, rp->bmain, rp->scene, lay, 0); // 0= dont use camera view
WM_job_main_thread_lock_release(rp->job);
/* do preprocessing like building raytree, shadows, volumes, SSS */
RE_Database_Preprocess(re);
/* conversion not completed, need to do it again */
if (!rstats->convertdone) {
if (render_view3d_is_valid(rp)) {
rp->engine->job_update_flag |= PR_UPDATE_DATABASE;
}
}
// printf("dbase update\n");
}
else {
// printf("dbase rotate\n");
RE_DataBase_IncrementalView(re, rp->viewmat, 0);
restore = 1;
}
RE_DataBase_ApplyWindow(re);
/* OK, can we enter render code? */
if (rstats->convertdone) {
bool first_time = true;
if (update_flag & PR_UPDATE_VIEW) {
ob_inst_update_flag |= RE_OBJECT_INSTANCES_UPDATE_VIEW;
}
RE_updateRenderInstances(re, ob_inst_update_flag);
for (;;) {
int pixel_size = BKE_render_preview_pixel_size(&rp->scene->r);
if (first_time == false) {
if (restore)
RE_DataBase_IncrementalView(re, rp->viewmat, 1);
2017-10-06 16:56:41 +11:00
rp->resolution_divider = MAX2(rp->resolution_divider / 2, pixel_size);
*do_update = 1;
render_update_resolution(re, rp, use_border, &cliprct);
RE_DataBase_IncrementalView(re, rp->viewmat, 0);
RE_DataBase_ApplyWindow(re);
restore = 1;
}
else {
render_update_resolution(re, rp, use_border, &cliprct);
}
RE_TileProcessor(re);
first_time = false;
if (*stop || rp->resolution_divider == pixel_size) {
break;
}
}
/* always rotate back */
if (restore)
RE_DataBase_IncrementalView(re, rp->viewmat, 1);
}
}
static void render_view3d_free(void *customdata)
{
RenderPreview *rp = customdata;
MEM_freeN(rp);
}
static bool render_view3d_flag_changed(RenderEngine *engine, const bContext *C)
{
RegionView3D *rv3d = CTX_wm_region_view3d(C);
View3D *v3d = CTX_wm_view3d(C);
ARegion *ar = CTX_wm_region(C);
Scene *scene = CTX_data_scene(C);
Render *re;
rctf viewplane;
rcti disprect;
float clipsta, clipend;
bool orth;
int job_update_flag = 0;
char name[32];
/* ensure render engine exists */
re = engine->re;
if (!re) {
sprintf(name, "View3dPreview %p", (void *)ar);
re = engine->re = RE_GetRender(name);
if (!re)
re = engine->re = RE_NewRender(name);
engine->update_flag |= RE_ENGINE_UPDATE_DATABASE;
}
/* check update_flag */
if (engine->update_flag & RE_ENGINE_UPDATE_MA)
job_update_flag |= PR_UPDATE_MATERIAL;
if (engine->update_flag & RE_ENGINE_UPDATE_OTHER)
job_update_flag |= PR_UPDATE_MATERIAL;
if (engine->update_flag & RE_ENGINE_UPDATE_DATABASE) {
job_update_flag |= PR_UPDATE_DATABASE;
/* load editmesh */
if (scene->obedit)
ED_object_editmode_load(scene->obedit);
}
engine->update_flag = 0;
/* check if viewport changed */
if (engine->last_winx != ar->winx || engine->last_winy != ar->winy) {
engine->last_winx = ar->winx;
engine->last_winy = ar->winy;
job_update_flag |= PR_UPDATE_RENDERSIZE;
}
if (compare_m4m4(engine->last_viewmat, rv3d->viewmat, 0.00001f) == 0) {
copy_m4_m4(engine->last_viewmat, rv3d->viewmat);
job_update_flag |= PR_UPDATE_VIEW;
}
render_view3d_get_rects(ar, v3d, rv3d, &viewplane, engine, &clipsta, &clipend, NULL, &orth);
if (BLI_rctf_compare(&viewplane, &engine->last_viewplane, 0.00001f) == 0) {
engine->last_viewplane = viewplane;
job_update_flag |= PR_UPDATE_VIEW;
}
render_view3d_disprect(scene, ar, v3d, rv3d, &disprect);
if (BLI_rcti_compare(&disprect, &engine->last_disprect) == 0) {
engine->last_disprect = disprect;
job_update_flag |= PR_UPDATE_RENDERSIZE;
}
/* any changes? go ahead and rerender */
if (job_update_flag) {
engine->job_update_flag |= job_update_flag;
return true;
}
return false;
}
static void render_view3d_do(RenderEngine *engine, const bContext *C)
{
wmJob *wm_job;
RenderPreview *rp;
Scene *scene = CTX_data_scene(C);
ARegion *ar = CTX_wm_region(C);
int width = ar->winx, height = ar->winy;
int divider = BKE_render_preview_pixel_size(&scene->r);
int resolution_threshold = scene->r.preview_start_resolution *
scene->r.preview_start_resolution;
if (CTX_wm_window(C) == NULL)
return;
if (!render_view3d_flag_changed(engine, C))
return;
wm_job = WM_jobs_get(CTX_wm_manager(C), CTX_wm_window(C), CTX_wm_region(C), "Render Preview",
2013-04-17 01:16:23 +00:00
WM_JOB_EXCL_RENDER, WM_JOB_TYPE_RENDER_PREVIEW);
rp = MEM_callocN(sizeof(RenderPreview), "render preview");
rp->job = wm_job;
while (width * height > resolution_threshold) {
width = max_ii(1, width / 2);
height = max_ii(1, height / 2);
divider *= 2;
}
/* customdata for preview thread */
rp->scene = scene;
rp->engine = engine;
rp->sa = CTX_wm_area(C);
rp->ar = CTX_wm_region(C);
rp->v3d = rp->sa->spacedata.first;
rp->rv3d = CTX_wm_region_view3d(C);
rp->bmain = CTX_data_main(C);
rp->resolution_divider = divider;
rp->start_resolution_divider = divider;
rp->has_freestyle = (scene->r.mode & R_EDGE_FRS) != 0;
copy_m4_m4(rp->viewmat, rp->rv3d->viewmat);
/* clear info text */
engine->text[0] = '\0';
/* setup job */
WM_jobs_customdata_set(wm_job, rp, render_view3d_free);
2013-04-17 01:16:23 +00:00
WM_jobs_timer(wm_job, 0.1, NC_SPACE | ND_SPACE_VIEW3D, NC_SPACE | ND_SPACE_VIEW3D);
WM_jobs_callbacks(wm_job, render_view3d_startjob, NULL, NULL, NULL);
WM_jobs_start(CTX_wm_manager(C), wm_job);
engine->flag &= ~RE_ENGINE_DO_UPDATE;
}
2015-02-11 18:38:41 +11:00
/* callback for render engine, on changes */
void render_view3d_update(RenderEngine *engine, const bContext *C)
{
/* this shouldn't be needed and causes too many database rebuilds, but we
2016-07-16 17:48:57 +10:00
* aren't actually tracking updates for all relevant datablocks so this is
* a catch-all for updates */
engine->update_flag |= RE_ENGINE_UPDATE_DATABASE;
render_view3d_do(engine, C);
}
void render_view3d_draw(RenderEngine *engine, const bContext *C)
{
Render *re = engine->re;
RenderResult rres;
char name[32];
render_view3d_do(engine, C);
if (re == NULL) {
sprintf(name, "View3dPreview %p", (void *)CTX_wm_region(C));
re = RE_GetRender(name);
if (re == NULL) return;
}
Multi-View and Stereo 3D Official Documentation: http://www.blender.org/manual/render/workflows/multiview.html Implemented Features ==================== Builtin Stereo Camera * Convergence Mode * Interocular Distance * Convergence Distance * Pivot Mode Viewport * Cameras * Plane * Volume Compositor * View Switch Node * Image Node Multi-View OpenEXR support Sequencer * Image/Movie Strips 'Use Multiview' UV/Image Editor * Option to see Multi-View images in Stereo-3D or its individual images * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images I/O * Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images Scene Render Views * Ability to have an arbitrary number of views in the scene Missing Bits ============ First rule of Multi-View bug report: If something is not working as it should *when Views is off* this is a severe bug, do mention this in the report. Second rule is, if something works *when Views is off* but doesn't (or crashes) when *Views is on*, this is a important bug. Do mention this in the report. Everything else is likely small todos, and may wait until we are sure none of the above is happening. Apart from that there are those known issues: * Compositor Image Node poorly working for Multi-View OpenEXR (this was working prefectly before the 'Use Multi-View' functionality) * Selecting camera from Multi-View when looking from camera is problematic * Animation Playback (ctrl+F11) doesn't support stereo formats * Wrong filepath when trying to play back animated scene * Viewport Rendering doesn't support Multi-View * Overscan Rendering * Fullscreen display modes need to warn the user * Object copy should be aware of views suffix Acknowledgments =============== * Francesco Siddi for the help with the original feature specs and design * Brecht Van Lommel for the original review of the code and design early on * Blender Foundation for the Development Fund to support the project wrap up Final patch reviewers: * Antony Riakiotakis (psy-fi) * Campbell Barton (ideasman42) * Julian Eisel (Severin) * Sergey Sharybin (nazgul) * Thomas Dinged (dingto) Code contributors of the original branch in github: * Alexey Akishin * Gabriel Caraballo
2015-04-06 10:40:12 -03:00
/* Viewport render preview doesn't support multiview, view hardcoded to 0 */
RE_AcquireResultImage(re, &rres, 0);
if (rres.rectf) {
RegionView3D *rv3d = CTX_wm_region_view3d(C);
View3D *v3d = CTX_wm_view3d(C);
Scene *scene = CTX_data_scene(C);
ARegion *ar = CTX_wm_region(C);
bool force_fallback = false;
bool need_fallback = true;
float dither = scene->r.dither_intensity;
float scale_x, scale_y;
rcti clip_rect;
int xof, yof;
if (render_view3d_disprect(scene, ar, v3d, rv3d, &clip_rect)) {
scale_x = (float) BLI_rcti_size_x(&clip_rect) / rres.rectx;
scale_y = (float) BLI_rcti_size_y(&clip_rect) / rres.recty;
xof = clip_rect.xmin;
yof = clip_rect.ymin;
}
else {
scale_x = (float) ar->winx / rres.rectx;
scale_y = (float) ar->winy / rres.recty;
xof = rres.xof;
yof = rres.yof;
}
/* If user decided not to use GLSL, fallback to glaDrawPixelsAuto */
force_fallback |= (U.image_draw_method != IMAGE_DRAW_METHOD_GLSL);
/* Try using GLSL display transform. */
if (force_fallback == false) {
if (IMB_colormanagement_setup_glsl_draw(&scene->view_settings, &scene->display_settings, dither, true)) {
glEnable(GL_BLEND);
glColor4f(1.0f, 1.0f, 1.0f, 1.0f);
glPixelZoom(scale_x, scale_y);
glaDrawPixelsTex(xof, yof, rres.rectx, rres.recty,
GL_RGBA, GL_FLOAT, GL_NEAREST, rres.rectf);
glPixelZoom(1.0f, 1.0f);
glDisable(GL_BLEND);
IMB_colormanagement_finish_glsl_draw();
need_fallback = false;
}
}
/* If GLSL failed, use old-school CPU-based transform. */
if (need_fallback) {
unsigned char *display_buffer = MEM_mallocN(4 * rres.rectx * rres.recty * sizeof(char),
"render_view3d_draw");
IMB_colormanagement_buffer_make_display_space(rres.rectf, display_buffer, rres.rectx, rres.recty,
4, dither, &scene->view_settings, &scene->display_settings);
glEnable(GL_BLEND);
glColor4f(1.0f, 1.0f, 1.0f, 1.0f);
glPixelZoom(scale_x, scale_y);
glaDrawPixelsAuto(xof, yof, rres.rectx, rres.recty,
GL_RGBA, GL_UNSIGNED_BYTE,
GL_NEAREST, display_buffer);
glPixelZoom(1.0f, 1.0f);
glDisable(GL_BLEND);
MEM_freeN(display_buffer);
}
}
RE_ReleaseResultImage(re);
}
void ED_viewport_render_kill_jobs(wmWindowManager *wm,
Main *bmain,
bool free_database)
{
bScreen *sc;
ScrArea *sa;
ARegion *ar;
if (!wm)
return;
/* kill all actively running jobs */
WM_jobs_kill(wm, NULL, render_view3d_startjob);
/* loop over 3D view render engines */
for (sc = bmain->screen.first; sc; sc = sc->id.next) {
for (sa = sc->areabase.first; sa; sa = sa->next) {
if (sa->spacetype != SPACE_VIEW3D)
continue;
for (ar = sa->regionbase.first; ar; ar = ar->next) {
RegionView3D *rv3d;
if (ar->regiontype != RGN_TYPE_WINDOW)
continue;
rv3d = ar->regiondata;
if (rv3d->render_engine) {
/* free render database now before we change data, because
* RE_Database_Free will also loop over blender data */
if (free_database) {
char name[32];
Render *re;
sprintf(name, "View3dPreview %p", (void *)ar);
re = RE_GetRender(name);
if (re)
RE_Database_Free(re);
/* tag render engine to update entire database */
rv3d->render_engine->update_flag |= RE_ENGINE_UPDATE_DATABASE;
}
else {
/* quick shader update */
rv3d->render_engine->update_flag |= RE_ENGINE_UPDATE_MA;
}
}
}
}
}
}
Scene *ED_render_job_get_scene(const bContext *C)
{
wmWindowManager *wm = CTX_wm_manager(C);
RenderJob *rj = (RenderJob *)WM_jobs_customdata_from_type(wm, WM_JOB_TYPE_RENDER);
if (rj)
return rj->scene;
return NULL;
}
Scene *ED_render_job_get_current_scene(const bContext *C)
{
wmWindowManager *wm = CTX_wm_manager(C);
RenderJob *rj = (RenderJob *)WM_jobs_customdata_from_type(wm, WM_JOB_TYPE_RENDER);
if (rj) {
return rj->current_scene;
}
return NULL;
}
/* Motion blur curve preset */
static int render_shutter_curve_preset_exec(bContext *C, wmOperator *op)
{
Scene *scene = CTX_data_scene(C);
CurveMapping *mblur_shutter_curve = &scene->r.mblur_shutter_curve;
CurveMap *cm = mblur_shutter_curve->cm;
int preset = RNA_enum_get(op->ptr, "shape");
cm->flag &= ~CUMA_EXTEND_EXTRAPOLATE;
mblur_shutter_curve->preset = preset;
curvemap_reset(cm,
&mblur_shutter_curve->clipr,
mblur_shutter_curve->preset,
CURVEMAP_SLOPE_POS_NEG);
curvemapping_changed(mblur_shutter_curve, false);
return OPERATOR_FINISHED;
}
void RENDER_OT_shutter_curve_preset(wmOperatorType *ot)
{
PropertyRNA *prop;
static const EnumPropertyItem prop_shape_items[] = {
{CURVE_PRESET_SHARP, "SHARP", 0, "Sharp", ""},
{CURVE_PRESET_SMOOTH, "SMOOTH", 0, "Smooth", ""},
{CURVE_PRESET_MAX, "MAX", 0, "Max", ""},
{CURVE_PRESET_LINE, "LINE", 0, "Line", ""},
{CURVE_PRESET_ROUND, "ROUND", 0, "Round", ""},
{CURVE_PRESET_ROOT, "ROOT", 0, "Root", ""},
{0, NULL, 0, NULL, NULL}};
ot->name = "Shutter Curve Preset";
ot->description = "Set shutter curve";
ot->idname = "RENDER_OT_shutter_curve_preset";
ot->exec = render_shutter_curve_preset_exec;
prop = RNA_def_enum(ot->srna, "shape", prop_shape_items, CURVE_PRESET_SMOOTH, "Mode", "");
RNA_def_property_translation_context(prop, BLT_I18NCONTEXT_ID_CURVE); /* Abusing id_curve :/ */
}