Issue was caused by the wrong usage of OCIO GLSL binding API. To make it
work properly on pre-GLSL-1.3 drivers shader is to be enabled after the
texture is binded to the opengl context. Otherwise it wouldn't know the
proper texture size.
This is actually a regression in 2.70 and to be ported to 'a'.
Z, Index, normal, UV and vector passes are only affected by surfaces with alpha
transparency equal to or higher than this threshold. With value 0.0 the first
surface hit will always write to these passes, regardless of transparency. With
higher values surfaces that are mostly transparent can be skipped until an opaque
surface is encountered.
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
It's a simple estimate, not very precise but that isn't really possible always.
For progressive render it will become more accurate the longer you render.
Reviewed By: brecht
Differential Revision: http://developer.blender.org/D67
except for curves, that's still missing from the OpenColorIO GLSL shader.
The pixels are stored in a half float texture, converterd from full float with
native GPU instructions and SIMD on the CPU, so it should be pretty quick.
Using a GLSL shader is useful for GPU render because it avoids a copy through
CPU memory.
* Render Passes are now available for Subsurface Scattering (Direct, Indirect and Color pass).
This is part of my GSoC project, SVN merge of r58587, r58828 and r58835.
a maximized Blender window in Ubuntu Unity. The window size would slightly change
as part of the unminimize effect.
Now cycles waits 0.2 seconds before restarting rendering after a viewport resize,
also a bit less flickery when changing the 3d view size in Blender itself.
panel now has an option to specify how to use them. There's three options:
* Use: render layer samples override scene samples
* Bounded: bound render layer samples by scene samples
* Ignore: ignore render layer sample settings
per render layer samples in addition to the progress bar.
Also fixed job progress bar not working at all on high DPI / retina, was so small
the actual progress was not visible.
- pass string size to BLI_timestr() to avoid possible buffer overrun.
- quiet warning for mingw.
- include guards for windows utf conversion funcs.
- fix for mistage in edge-angle-selection check.
- some style cleanup.
before this only active scene would be rendered with border.
When do_render_fields_blur_3d() is finished, it'll modify
render's display rect so it'll correspond bordered render
result placed on black backgrund. Actual border is stored
nowhere, which makes it only way to re-calculate disprect
for all other renders used in compo based on source. Not
so big deal actually.
Also needed to modify Cycles a bit, because before this
patch it used border settings from scene being rendered.
Now made it so render data is passing to external engines.
Using a property inside RenderEngine structure for this.
Not best ever design for passing render data, but this
would prevent API breakage. So now external engines could
access engine.render to access active rendering settings.
Reviewed by Brecht, thanks!
Yes, again.
There's some t within which reset is not allowed. This is so no reset happens
too often for performance issues. If camera changes too often, some reset could
be missed because of this timeout.
For now tag engine for update, which will update viewport from blender side.
Proper solution could be to detect such a changes from blender side and tag
cycles for refresh instead of trying to detect changes form cycles, but that's
for later.
Issue was caused by rare cases when camera move happens just after
last sample was finished, this would lead to missing delay reset
because render cycle will go to pause_cond.wait(). No reset will
happen at this point because of some kind of optimization which
checks whether camera is tagged for update and wouldn't do reset
in this case.
Talked to Brecht and seems this optimization is not actually needed
and removing it will solve issue with frozen preview.
Issue was caused by cycles trying to find builtin images in a main
database and in case of preview render images are not in database,
they're just referenced by shader node tree.
Now builtin images in cycles have got void* pointer to store data
needed to load builtin images.
In case ob blender session, this pointer will store pointer from
PointerRNA for image datablock and used later to construct Image
class based on this pointer.
This also saves database lookup for final render which is nice :)
Reviewed by Brecht.
Patch by Sergey, .blend by Thomas and some further tweaks by me.
Still to solve later: allow external engines to specify own preview .blend, for
now the code here is doing too much magic hacking on the preview scene still.
This adds support of movie textures for Cycles rendering.
Uses the same builtin images routines as packed/generated images,
but with some extra non-rna hookups from blender_session side.
Basically, it's not so clear how to give access to video frames
via C++ RNA -- it'll require exposing ImBuf to API, doing some
threading locks and so. Ended up adding two more functions which
are actually bad level call, but don't consider it's so much bad
-- we have few bad calls already, which are actually related.
Changed a bit how builtin images names are passing to image
manager. Now it's not just an ID datablock name, but also a frame
number concatenated via '@' character, which makes itpossible to
easily know frame number to be used for movie images, without
adding extra descriptors to image manager.
Decoding of builtin name is a bit slower now, but it should be
still nothing in comparison with rendering complexity.
Also exposed image user's frame_current to python API, which
is needed to get absolute frame number of movie from node's
image user.
P.S. Generated/packed images are also using bad level call but
only does it to make things more clear here. Either all images
are using C++ RNA here or no images does. That's the most clear
for now.
This commit adds support of packed and generated images
for Cycles when using SVM backend. Movies are still not
supported. This changes also doesn't touch OSL which is
much less trivial to adopt for any images which are not
saved to disk.
Implementation details:
- When adding images to Image Manager is now possible
to mark image as builtin. Builtin images will bypass
OIIO loader and will use special loading callbacks.
- Callbacks are set by Blender Session and they're
using C++ RNA interface to obtain needed data (pixels,
dimensions, is_float flag).
- Image Manager assumes file path is used as reference
to a builtin images, but in fact currently image
datablock name is used for reference. This makes it
easy to find an image in BlendData database.
- Added some extra properties to Image RNA:
* channels, which denotes actual number of channels
in ImBuf. This is needed to treat image's pixels
correct (before it wasn't possible because API
used internal number of channels for pixels which
is in fact doesn't correlate with image depth)
* is_float, which is truth if image is stored in
float buffer of ImBuf.
- Implemented string lookup for C++ RNA collections
for cases there's no manual lookup function.
OSL is not supported because it used own image loading
and filtering routines and there's seems to be no API
to feed pre-loaded pixels directly to the library.
Think we'll either need to add some API to support
such kind of feeding or consider OSL does not have
support of packed images at all.
Movies are not supported at this moment because of lack
of RNA API to load specified frame. It's not difficult
to solve, just need to consider what to best here:
* Either write some general python interface for ImBuf
and use it via C++ API, or
* Write a PY API function which will return pixels for
given frame, or
* Use bad-level BKE_* call
Anyway, small steps, further improvements later.
Reviewed by Brecht, thanks!
This commit implements highlight of tiles which are being currently
rendered for both Blender Internal and Cycles (and should be possible
to use it for other external engines as well).
Couple of implementation details:
- Added one extra boolean flag to render engine which should be set
to truth if render engine wants to highlight tiles. If so, property
use_highlight_tiles should be set to True.
- Render Part's ready boolena was changed by status enum, which could
be NONE, IN_PROGRESS and READY. All render part with IN_PROGRESS
status will be highlighted in image editor.
- For external engines render part's status is filling in automatically.
Initially all render parts has got NONE status, then one external
engine acquire render result, corresponding part will change status
to IN_PROGRESS. As soon as render result is finished, corresponding
render part will change status to FINISHED
This should make it easy to highlight tiles for other engines as well.
Patch [#33445] - Experimental Cycles Hair Rendering (CPU only)
This patch allows hair data to be exported to cycles and introduces a new line segment primitive to render with.
The UI appears under the particle tab and there is a new hair info node available.
It is only available under the experimental feature set and for cpu rendering.
This option enables keeping loaded images in the memory in-between
of rendering.
Implemented by keeping render engine alive for until Render structure
is being freed.
Cycles will free all data when render finishes, optionally keeping
image manager untouched. All shaders, meshes, objects will be
re-allocated next time rendering happens.
Cycles cession and scene will be re-created from scratch if render/
scene parameters were changed.
This will also allow to keep compiled OSL shaders in memory without
need to re-compile them again.
P.S. Performance panel could be cleaned up a bit, not so much happy
with it's vertical alignment currently but not sure how to make
it look better.
P.P.S. Currently the only way to free images from the device is to
disable Persistent Images option and start rendering.
This commit adds memory usage information while rendering.
It reports memory used by device, meaning:
- For CPU it'll report real memory consumption
- For GPU rendering it'll report GPU memory consumption, but it'll
also mean the same memory is used from host side.
This information displays information about memory requested by Cycles,
not memory really allocated on a device. Real memory usage might be
higher because of memory fragmentation or optimistic memory allocator.
There's really nothing we can do against this.
Also in contrast with blender internal's render cycles memory usage
does not include memory used by scene, only memory needed by cycles
itself will be displayed. So don't freak out if memory usage reported
by cycles would be much lower than blender internal's.
This commit also adds RenderEngine.update_memory_stats callback which
is used to tell memory consumption from external engine to blender.
This information is used to generate information line after rendering
is finished.
Added support of such features, as:
- Ability to call RNA functions using C++ classes
For example RenderEngine.tag_update
- Property setters (for scalars and arrays)
Used Qt/jQuery-like getters/setters style, meaning Class.prop() is a getter,
Class.prop(value) is a setter.
Still to come:
Collection functions are not currently registering inside a property
Meaning BlendData.meshes wouldn't be a subclass of BlendDataMeshes result
you'll need to explicitly create BlendDataMeshes for now instead of doing
BlendData.meshes.remove()
This makes it possible to do a border render inside a viewport even
when not looking through the camera.
Render border could be defined by Ctrl-B shortcut (works for both
camera render border and viewport render border).
Camera render border could still be defined using Shift-B (so no
muscule memory would be broken). Currently used a special flag of
operator to do this, otherwise you'll need to either two operators
with different poll callback or it could go into conflict with a
border zoom,
Border render of a viewport could be enabled/disabled in View
panel using "Render Border" option.
Just makes progressive refine :)
This means the whole image would be refined gradually using as much
threads as it's set in performance settings. Having enough tiles is
required to have this option working as it's expected.
Technically it's implemented by repeatedly computing next sample for
all the tiles before switching to next sample.
This works around 7-12% slower than regular tile-based rendering, so
use this option only if you really need it.
This commit also fixes progressive update of image when Save Buffers
option is enabled.
And one more thing this commit fixes is handling display buffer with
Save Buffers option enabled. If this option is enabled image buffer
wouldn't have neither byte nor float buffer until image is fully
rendered which could backfire in missing image while rendering in
cases color management cache became full.
This issue solved by allocating byte buffer for image buffer from
tile update callback.
Patch was reviewed by Brecht. He also made some minor edits to
original version to patch. Thanks, man!
When the scene is updated Cycles resets the renderer device, cancelling
all existing tasks. The main thread would wait for all running tasks to
finish before continuing. This is ok when tasks can actually cancel in a
timely fashion. For OSL however, this does not work, since the OSL
shader group optimization takes quite a bit of time and can not be
easily be cancelled once running (on my crappy machine in full debug
mode: ~0.12 seconds for simple node trees). This would lead to very
laggy UI behavior and make it difficult to accurately control elements
such as sliders.
This patch removes the wait condition from the device->task_cancel
method. Instead it just sets the do_cancel flag and returns. To avoid
backlog in the task pool of the device it will return early from the
BlenderSession::sync function while the reset is going on (tested in
Session::resetting). Once all existing tasks have finished the do_cancel
flag is finally cleared again (checked in TaskPool::num_decrease).
Care has to be taken to avoid race conditions on the do_cancel flag,
since it can now be modified outside the TaskPool::cancel function
itself. For this purpose the scope of the TaskPool::num_mutex locks has
been extended, in most cases the mutex is now locked by the TaskPool
itself before calling TaskScheduler methods, instead of only locking
inside the num_increase/num_decrease functions themselves. The only
occurrence of a lock outside of the TaskPool methods is in
TaskScheduler::thread_run.
This patch is most useful in combination with the OSL renderer mode, so
it can probably wait until after the 2.64 release. SVM tasks tend to be
cancelled quickly, so the effect is less noticeable.
Regular rendering now works tiled, and supports save buffers to save memory
during render and cache render results.
Brick texture node by Thomas.
http://wiki.blender.org/index.php/Doc:2.6/Manual/Render/Cycles/Nodes/Textures#Brick_Texture
Image texture Blended Box Mapping.
http://wiki.blender.org/index.php/Doc:2.6/Manual/Render/Cycles/Nodes/Textures#Image_Texturehttp://mango.blender.org/production/blended_box/
Various bug fixes by Sergey and Campbell.
* Fix for reading freed memory in some node setups.
* Fix incorrect memory read when synchronizing mesh motion.
* Fix crash appearing when direct light usage is different on different layers.
* Fix for vector pass gives wrong result in some circumstances.
* Fix for wrong resolution used for rendering Render Layer node.
* Option to cancel rendering when doing initial synchronization.
* No more texture limit when using CPU render.
* Many fixes for new tiled rendering.
