FFmpeg expects Blender to feed it pixels in the output pixel format. If
the output pixel format is different than Blender's RGBA, a conversion
is needed (via FFmpeg's `sws_scale()` function). There were a few issues
with this conversion (and surrounding code) that are fixed in this
commit:
- When conversion was necessary a temporary buffer was allocated and
deallocated for every frame. This is now allocated once and re-used.
- Copying data to the buffer was done byte-for-byte. On little-endian
machines it is now done line-by-line using `memcpy` for a little speedup.
- The decision whether pixel format conversion is necessary is now
correctly done based on the pixel format Blender is actually using.
- The pixel format of the buffer sent to FFmpeg is no longer hard-coded
incorrectly to a fixed pixel format, but uses the actual output pixel
format. This is fixes T53058 properly, making RGB QTRLE export possible.
- I added some comments to make it clear which pixel format is referred
to (either Blender's internal format or the FFmpeg output format).
Combined these improvements not only correct a bug (T53058) but also
results in approximately 5% speed improvement (tested with a 117-frame
shot from Spring, loaded as PNGs in the VSE, encoding to h.264 with
preset 'realtime').
Reviewed By: brecht, sergey
Differential Revision: https://developer.blender.org/D5174
This adds a space between a value and its short unit name except for foot, inch, degree, arcminute, arcsecond
Differential Revision: https://developer.blender.org/D5051
Reviewed by Brecht Van Lommel
Was a nice looking invention, but is obviously wrong.
Is likely no function changes, since -false is like -0
which is 0 and which false. And -true is like -<non-zero> which
is non-zero as well.
Spotted by Sybren, thanks!
This change ensures that operators which needs access to evaluated data
first makes sure there is a dependency graph.
Other accesses to the dependency graph made it more explicit about
whether they just need a valid dependency graph pointer or whether they
expect the graph to be already evaluated.
This replaces OPTYPE_USE_EVAL_DATA which is now removed.
Some general rules about usage of accessors:
- Drawing is expected to happen from a fully evaluated dependency graph.
There is now a function to access it, which will in the future control
that dependency graph is actually evaluated.
This check is not yet done because there are some things to be taken
care about first: for example, post-update hooks might leave scene in
a state where something is still tagged for update.
- All operators which needs to access evaluated state must use
CTX_data_ensure_evaluated_depsgraph().
This function replaces OPTYPE_USE_EVAL_DATA.
The call is generally to be done in the very beginning of the
operator, prior other logic (unless this is some comprehensive
operator which might or might not need access to an evaluated state).
This call is never to be used from a loop.
If some utility function requires evaluated state of dependency graph
the graph is to be passed as an explicit argument. This way it is
clear that no evaluation happens in a loop or something like this.
- All cases which needs to know dependency graph pointer, but which
doesn't want to actually evaluate it can use old-style function
CTX_data_depsgraph_pointer(), assuming that underlying code will
ensure dependency graph is evaluated prior to accessing it.
- The new functions are replacing OPTYPE_USE_EVAL_DATA, so now it is
explicit and local about where dependency graph is being ensured.
This commit also contains some fixes of wrong usage of evaluation
functions on original objects. Ideally should be split out, but in
reality with all the APIs being renamed is quite tricky.
Fixes T67454: Blender crash on rapid undo and select
Speculation here is that sometimes undo and selection operators are
sometimes handled in the same event loop iteration, which leaves
non-evaluated dependency graph.
Fixes T67973: Crash on Fix Deforms operator
Fixes T67902: Crash when undo a loop cut
Reviewers: brecht
Reviewed By: brecht
Subscribers: lichtwerk
Maniphest Tasks: T67454
Differential Revision: https://developer.blender.org/D5343
Unlike `fill_texpaint_slots_recursive`, `rna_Material_active_paint_texture_index_update` did not search for texture nodes that are inside node groups.
Reviewers: sergey, psy-fi, zeddb, brecht
Subscribers: brecht
Differential Revision: https://developer.blender.org/D5338
font_to_curve code was not handling properly the case where it has nop
object to check for mat indices validity. Check should just not happen
then, not reset mat indices of chars to default 0 value.
This fixes an incompatibility with Visual Studio 2019 introduced in
631d5026c7. It is likely caused by using
`# ifdef` inside the use of the `ELEM()` macro.
Use explicit boolean flag to indicate whether flush to original data
is needed or not. Makes it possible to avoid confusion on whether an
evaluated or any depsgraph can be passed to the API.
Allows to remove depsgraph from bAnimContext as well.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5379
Simplify preferences by removing the ability to load them from
either the startup.blend or userpref.blend.
Also simplifies updating default preferences by moving
them to a struct definition.
This patch introduces the same presets that are used for proportional editing in the brush falloff menu. The user can select any of these presets or use the regular custom falloff curve. The presets are hardcoded formulas, so the falloff curve is not used when they are active.
This change improves the general feeling of the brushes and it is more convenient and simpler to use. The CUSTOM curve option should now be used in the case that an unusual deformation is needed, in other cases, the hardcoded curve presets should be the default.
The smooth curve presets is a must in the grab brush, as it fixes the deformation issue with the current custom curve setting. The user may try to adjust the deformation by tweaking the curve, but it is nearly impossible to replicate this desired behavior.
{F7636217}
Other brushes that are included in the sculpt branch also rely on this as they need specific hardcoded falloffs to produce the desired effect.
Reviewers: brecht, billreynish
Reviewed By: brecht
Subscribers: JulienKaspar
Differential Revision: https://developer.blender.org/D5367
This function makes it possible to clear/remove/nuke all the geometry in
a mesh, allowing functions like `Mesh.from_python()` to construct a new
mesh in its place. Without this function, code like in T67627 have to
allocate a new Mesh datablock, fill that, and swap out the old Mesh for
the new one. This introduces issues when exporting, as the new mesh
could be seen by an exporter as unrelated to the old one.
Shape keys are not freed by this function. Freeing those would require
tagging the depsgraph for relations update, which is an expensive
operation. They should be removed explicitly if necessary.
Material slots are also not cleared by this function, in the same way
that they are not cleared when manually removing all geometry from a
mesh.
The `BKE_mesh_clear_geometry()` function is available in Python as
`mesh.clear_geometry()`.
Reviewed by: mont29, brecht
Differential Revision: https://developer.blender.org/D5373
This commit adds support for the WebM container. Previously we only
supported the WebM/VP9 video codec, but still required that it was
stored in a Matroska, MP4, or other compatible container format.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D5156
The VP9 video codec supports writing alpha values; now this is available
in Blender too.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D5156
Simplified `BKE_ffmpeg_alpha_channel_is_supported()` to use `ELEM()`
instead of a row consecutive `if`-statements.
No functional changes.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D5156
Some ugly very low-level collection code was using the generic
LIB_TAG_DOIT tag... should never happen, that one is for rather
high-level code to use, core process shall use own tags.
This was caused by 2D stabilization trying to be smart and lower weight
of tracks which are too close to the rotation center. This was causing
algorithm to ignore a single track which was set to constant 1 weight and
used for rotation compensation.
It is quite tricky to quantify this change without having comprehensive
regression suit, so can only hope that initial intention is still working
as expected.
Sync back between proxy armature pose and linked one was not properly
clearing the posebone runtime data, ending up with shared memory between
both (direct cause of reported crash), and probably some other nasty
issues.
While this crash is not critical in itself, I would suggest to add this
fix to 2.80 branch, as it is super-safe and simple, and is probably
fixing some other un-reported issues?
Main issue in that report was that meshes generated from
`mesh_create_eval_final_view()` and the like need some
special freeing handling (as, among other things, they borrow and do not
own their potential editmesh data...).
Factorized that into a helper func also used by
`BKE_object_free_derived_caches()`.
Such reload can no longer happen directly and is to be done via dependency
graph.
Eventually, the movie cache will become shared across all copies of the
clip, but even then we still need to have dependency graph mechanism because
we need to update FFmpeg animation handle (which can not be shared across
the copies).
Need to copy changes from original armature to evaluated one prior to
BKE_pose_where_is(), otherwise the pose matrix in pose channels will
not be correct.
Can not use dependency graph update here, since we must not re-evaluate
possible dependencies of the changed armature.
Reviewers: brecht
Reviewed By: brecht
Maniphest Tasks: T67507
Differential Revision: https://developer.blender.org/D5331
Before this the message could be too generic which was simply saying
that console is to be checked. This isn't very useful in cases when
we know that reconstruction is failed because of lack of good keyframes
or failure of initial reconstruction if there is no enough parallax
on the selected keyframes.
Correct fix that doesn't cause T67217.
Temporarily removing the excluded undo step broke memfile-undo
since freeing the undo steps needs to access other steps in
the list to merge shared chunks, see: memfile_undosys_step_free.
Pass the exclude step as an argument instead.
We need a final subversion bump before release in case there are
remaining 'unversioned' code in our versionning callbacks...
Should have been done before first RC actually. ;)
To be ported to 2.80 release branch.
