| Age | Commit message (Collapse) | Author |
|---|
|
|
|
These often happen in external libraries that we can't fix. The leaks are
still printed in the logs.
|
|
|
|
|
|
|
|
|
|
Undo rewrite broke test_undo.text_editor_edit_mode_mix.
Relatively harmless since it's only out of sync by one action.
|
|
Thanks to @sybren for pointing out issue from 299ff91ea1fe5
|
|
This reduces the number of loops
|
|
If the macro was used in a loop where the stroke was removed, the macro failed.
|
|
This function will be used for new operators.
|
|
followup to rB8dd95abb2ff9 (which fixed this for the Compositor node),
turns out this was also wrong for the VSE modifier and in vertex color
operator.
- also adjust min/max for VSE modifier
- also guard against division by zero
Reviewers: brecht
Maniphest Tasks: T67808
Differential Revision: https://developer.blender.org/D5398
|
|
This commit adds the frame to deformStroke Callback as a preparation for new modifiers that will need this variable.
Actually, the existing modifiers are not using the frame.
|
|
|
|
|
|
wheel
`applyResize(...)` considers that `t->values` always represents a `ratio`.
But this is only `true` with the `MOUSEMOVE` event.
The solution proposed is to never change `t->values`.
The result of the final transformation is now written to `t->values_final`.
Reviewers: campbellbarton
Differential Revision: https://developer.blender.org/D5212
|
|
|
|
The right thing is to be occluded by the visible mesh (in the case with modifiers).
|
|
|
|
|
|
|
|
|
|
Due to current uses, this does not bring functional changes.
|
|
|
|
While these aren't currently used, its strange to have
extended versions of a function without the non-extended versions
it also avoids callers needing to add them back if they need -
or duplicating the same boiler plate when calling the '_ex' versions.
Reverts 11da3b132aff1, 11da3b132aff1, adds depsgraph argument
so the caller is responsible for the evaluation state.
|
|
While harmless, use flow control that always sets the variable.
|
|
While harmless, use flow control that always sets the variable.
|
|
|
|
Matches shortcuts from the image editor.
D5341 by @tintwotin
|
|
Fix T67279
Fix T59954
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D5368
|
|
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D5219
|
|
|
|
There were several problems in the old random calculation:
* Different result in the viewport and render.
* Noise "pop" in some frames.
* Random number was calculated every time the file was opened, so get different results.
Now, instead to calculate the random numbers when n number of frames changed, the random values are calculated using a unique seed by stroke.
Also, a new Seed parameter has been added and this adds more control in the noise generated. This value can be animated and get more variations.
Differential Revision: http://developer.blender.org/D5393
|
|
|
|
given, instead of ignoring it.
Brings behavior of those options in line with the VGroup editing
modifiers ones e.g.
|
|
The mesh can be freed by BKE_mesh_nomain_to_mesh(), so we need to get
the `ME_AUTOSMOOTH` flag before that call, and not after.
|
|
Meshroom writes two hierarchies to Alembic, one rooted at
`/mvgRoot/mvgCameras` and the other at `/mvgRoot/mvgCamerasUndefined`.
These paths have no schema definition, and thus are ignored by Blender.
The cameras themselves have those schemaless paths as parent, and have
their transforms marked as "inherited", e.g. relative to their parent.
As these cameras have no valid parent, there is no Blender object to use
to convert their local matrices to world matrices, and Blender just
decided to reset them to the unit matrix.
Now "inherited" transforms without a parent in Blender are interpreted
as world transforms. Reparenting those objects to a Blender object will
re-interpret the transforms as local to the parent again.
|
|
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
|
|
Blender session
The depth attachment of the framebuffer was missing.
|
|
This is a standard combination (VP9 video, Opus audio, in WebM container),
so it's nice to have as a preset.
|
|
Those are now supported for tooltips.
|
|
|
|
Consistent with many code editors, see: D5175
|
|
|
|
|
|
|
|
Reviewers: sergey
Maniphest Tasks: T68036
Differential Revision: https://developer.blender.org/D5391
|
|
size_t wasn't declared. While this could be resolved differently,
convention is to include BKE after BLI.
|
|
Convention is to use ex, not ext for extended
versions of a functions.
|
|
T68045 by @luzpaz
|
