Introduced with d527e3a6bd.
Cached values are tagged as dirty during the update step, this can cause
conflicts where we attempt to then flush then changes into the PBVH but
have not yet updated the mesh pointers and reinitialized them.
This commit forcibly initializes the underlying data to prevent such
cases from happening when flushing to the PBVH.
Pull Request: https://projects.blender.org/blender/blender/pulls/125396
This is a PR that is built on top of #122094 (thanks to @SietseB
for the initial work).
Adds the following properties and functions:
* `layer.frames`,
* `layer.frames.new(frame_number)`
* `layer.frames.remove(frame_number)`
* `layer.frames.copy(from, to)`
* `frame = layer.get_frame_at(frame_number)`
* `frame.drawing`
* `frame.frame_number`
* `frame.select`: the selection state of the keyframe in the dope sheet
* `frame.keyframe_type`
* `drawing = frame.drawing`
* `drawing.type` (`DRAWING`/`REFERENCE`)
* `drawing.user_count`: The number of keyframes that use this drawing
* `drawing.attributes`: attribute read/write access to the drawing data
To be able to access attributes on drawings, a new
`AttributeOwnerType::GreasePencilDrawing` is added.
The API in `BKE_attributes.h` is updated to handle this type.
In `rna_attributes.cc`, there is a new
`rna_def_attribute_group_grease_pencil_drawing` that defines the
attribute group. For this to work, it also defines its own rna
callback functions.
Pull Request: https://projects.blender.org/blender/blender/pulls/124787
This was just rather useless level of abstraction. I heard from other
devs that these helper classes caused confusion, so better to avoid
this.
Now the asset representation has all the needed bits to create its full
path, blend-file path and asset library relative path. In fact only the
asset library relative path needs to be stored to make all this
available, since the asset representation already stores a reference to
the asset library owning it, so the paths can be recreated easily.
When writing a layered Action to disk, take the F-Curves from the
first keyframe strip and write that as `action.curves` as well. This
will make older Blender versions see those curves and load them
properly.
Only the curves for the first slot are written this way. This means
that any legacy Action that was converted to a layered Action will be
loaded again properly by older Blender versions. Of course this is
limited to a single layer, single strip, and single slot -- once the
newer features are used, older versions of Blender will not be able to
see this extra data.
When an Action contains multiple slots, so with animation for multiple
distinct objects, the forward compatibility becomes a bit iffy. Older
versions of Blender will just see a legacy Action, with its legacy
semantics, and thus all objects that use that Action will receive the
exact same animation data. I don't think there's a way around this.
(_Unless we start breaking up Actions into an Action per slot, alter
the assignments, and then store metadata so that modern Blenders can
reassemble them. I do not think this is a good idea._)
Ref: #124714
Pull Request: https://projects.blender.org/blender/blender/pulls/125065
Instead just compute the offsets as necessary. This avoids the need
to keep them in sync as the BMesh changes, though it requires passing
a few more arguments in the dynamic topology remeshing code.
This log is owned by SculptSession, it's a bit misleading
to store a pointer to it in the sculpt BVH tree, and having
multiple mutable pointers to objects should generally be
avoided anyway. Now just pass it to the remeshing function
which is the oinly place it was needed anyway.
Replace the `BKE_scene_cursor` functions with member functions
of the `View3DCursor` DNA struct. This makes using the cursor's
transform simpler, especially in newer C++ code.
Pull Request: https://projects.blender.org/blender/blender/pulls/124903
There were two issues:
* The target layer name was not copied over
* The thickness setting was being divided by two when it
resulted in thinner strokes.
Rename leftover references to action 'bindings' to 'slot':
- Two comments, and
- bunch of `bind_` variable prefixes, renamed to `slot_`.
No functional changes.
Add `#ifdef WITH_ANIM_BAKLAVA` to the blend file reading/writing code,
so that the Action layers & slots are ignored when Blender is built
without experimental features.
This ensures that any loaded Action is just treated as 'legacy' (which
is the only kind of Action non-experimental Blender should have to deal
with), which will also properly deal with the forward compatible data
written by !125065.
This fix was committed on the `blender-v4.2-release` branch as
1b7485f20892523942752f81239807b2eab0f00b.
Pull Request: https://projects.blender.org/blender/blender/pulls/125068
All image saving mechanisms in Blender ignores the color format for EXR
images, including Render Pipeline, Save Operator, and File Output nodes.
To fix this, we first allow EXR images to be BW for flexibility. Then we
adjust the EXR saving code to take into account the required number of
channels. This is only done for single layer EXR images. Multi-layer EXR
images correctly ignores the option.
Pull Request: https://projects.blender.org/blender/blender/pulls/124807
Though it results in more duplication currently, splitting these
could help to facilitate further performance improvements here
in the future, and it avoids passing a bunch of useless arguments
for the multires case.
These were mostly getting in the way of refactoring this code.
If the referenced problems actually happen, there would be
more obvious ways to observe the issues anyway.
This applies the same change as the previous commit to the bounds
of every BVH node. The bounds calculation can be changed to use
the standard functions from the regular BVH deformation.
In a simple test this makes building the sculpt BVH 64% faster.
I observed a change from 762ms to 464ms for a 1.9m vertex mesh.
This simplifies the BVH build process and potentially improves
its performance when parts of the geometry is hidden. The method
used to calculate whether a node is fully hidden is slightly different
too, vertex indices are used instead of triangle indices and a triangle
to face map.
Introduced in d527e3a6bd.
The change to update PBVH normals before destroying the PBVH to fix
shading on duplicate meshes issues had the unfortunate side effect of
causing crashes for multires due to a race condition. This commit
simply avoids performing this flushing for multires to allow for
further iteration in the future instead of reverting the offending
commit entirely.
Pull Request: https://projects.blender.org/blender/blender/pulls/125268
Part of #118145.
These days we aren't really benefiting from making PBVH an opaque type.
As we remove its responsibilities to focus it on being a BVH tree and look
to improve performance with data-oriented design, that will only become
more true.
There are some other future developments the current header structure
makes difficult:
- Storing selections of nodes with `IndexMask` for simpler iteration, etc.
- Specialization of node type for each PBVH type
- Reducing overhead of access to node data as nodes get smaller
- General C++ cleanliness and consistency
This PR moves `PBVH` to `blender::bke::pbvh::Tree` and moves `PBVHNode`
to `blender::bke::pbvh::Node`. Both are classes visible to elsewhere in Blender
but with private data fields.
The difficult part about the change is that we're in the middle of a transition
removing data from PBVH. Rather than making some data truly private I
chose to just give it the `_` suffix, since it will ideally be removed later.
Other things should be class methods or implemented as part of friend
classes. But the "fake" private status is much simpler for now and avoids
increasing the scope of this PR too much. Though that's a bit ugly, there's a
straightforward way to resolve these issues-- it just looks like the sort of
inconsistency you'd expect in the middle of a large refactor.
Pull Request: https://projects.blender.org/blender/blender/pulls/124919
In the case where the last layer is removed and all the drawings
have zero users, in the `remove_drawings_with_no_users` function
the `find_next_swap_index` lambda would return `false` on the
first call. Both `first_unused_drawing` and `last_used_drawing`
are `0` in this case. This meant that the `drawings_to_remove`
index mask would exclude the first drawing (because
`last_used_drawing` is the index of the first drawing) and not
remove it as it should.
To fix this, we check if `first_unused_drawing` is greater than zero.
If it is not, then we know all the drawings have to be removed.
Otherwise we only remove the drawings after
`last_used_drawing + 1`.
Pull Request: https://projects.blender.org/blender/blender/pulls/125318
This patch adds support for multi-pass compositing for EEVEE. This is
done by copying the passes used by the compositor node tree to the DRW
view data, which can then be accessed by the viewport compositor.
The viewport compositor will fallback to the viewport texture or an
invalid output of the passes were not initialized, this is currently the
case for any render engine that is not EEVEE.
A future optimization that we can do is eliminate the film pass copy
shaders and only copy the data that EEVEE rendered, which can be a
subset of the viewport for border rendering. This is not done at the
moment because not all engines support passes at the moment, so the
compositor expects full viewport passes.
Depends on: #123685, #123817, #123815.
Pull Request: https://projects.blender.org/blender/blender/pulls/123378
Regression from [0], drivers were tagged as being disabled with a flag
that was never cleared. Causing the label to be displayed for files
where the expressions were enabled and in use.
Resolve by clearing this flag on file load and when re-compiling
expressions - since an expressions block flag may be cleared if it
becomes a simple expression.
[0]: 1a8053939b
When attempting to replace `AttributeOwner` with `AttributeAccessor`, we ran into
the problem that the active attribute index is not stored on the `CurvesGeometry`.
The index can be moved between the two places without breaking forward- or backward-compatibility.
Pull Request: https://projects.blender.org/blender/blender/pulls/125059
Add an explicit mask in `IDAddOperations` for flags that are inherited
by default for dependencies of explicitely added data.
And refactor handling of per-id-usage flags returned by the
`dependencies_filter_cb` callback to also have a mask value defined in
`IDAddOperations`, as this is clearer and easier to maintain than a
constexpr hiddin in implementation code.
This commit changes the keying code to deselect keyframes when inserting new keys.
This has been discussed in the Animation & Rigging module meeting [1].
There is also an RCS post about that [2].
Doing this brings key creation in line with object creation,
where only the newly created object is selected.
There has been a previous attempt [3] to do a similar thing.
### Changes
When inserting keys by pressing `I` in the viewport or choosing a keying set,
all keys of the `Action` get deselected before inserting new keys.
New keys are selected by default.
Python RNA functions are **NOT** affected, meaning addons using
those functions will not deselect any keys by default.
The developer has to choose to do so.
To make that easier, there is a new RNA function on the action
`deselect_keys`
[1]: https://devtalk.blender.org/t/2024-05-02-animation-rigging-module-meeting/34493#patches-review-decision-time-5
[2]: https://blender.community/c/rightclickselect/K0hbbc
[3]: https://archive.blender.org/developer/D11623
Pull Request: https://projects.blender.org/blender/blender/pulls/121908
This is used by ID copy/paste code to detect which IDs from a copy/paste
buffer can be used as paste source.
Also slightly refactor implementation to handle both fake/extra ID user,
and clipboard mark in the same utils function.
`DUPLICATE_DEPENDENCIES` was wrongly used instead of `ADD_DEPENDENCIES`,
the former should be considered a non-modifiable option for a whole
'add' operation. But the per-ID-usage callback should be able to force a
dependency to be added or cleared.
There is no reason to deal with layer collections resync in non-main
case, deletion code should never trigger it anyway.
This will avoid quite a lot of the non-main-thread calls to
`BKE_layer_collection_resync_forbid` (detected and 'fixed' in previous
commit).
This commit allows re-entrant calls to
`BKE_layer_collection_resync_forbid` and
`BKE_layer_collection_resync_allow`, such that higher-level code can
block layer collections updates while calling lower-level code that does
the same thing.
E.g. `BKE_id_delete` blocks and then re-allows such updates. However,
such code can also be called from other pieces of code that also need
to prevent these updates in their own, broader scope.
In addition, realized that this call had concurrency issues since BKE
code using it is called from non-main thread too (e.g. from depsgraph
evaluation).
So made the counter atomic to ensure all calls to
`BKE_layer_collection_resync_forbid` is matched by a call to
`BKE_layer_collection_resync_allow`. Also added a cap to max amount of
're-entrant' calls (has to be quite high, due to potential concurrent
calls too).
This is required to fix#124777.
This fixes a few issues and clear up some confusion
in the code.
Note that this changes the behavior of render region;
they now reduce the internal render size. This is
matching the new design documentation.
- Data passes have correct accumulation.
- Adhere to naming conventions for extents, film and render pixels.
- Jitter over final pixels first before doing random sampling
in order to speed up convergence.
- Ensure enough sample to cover at least all the film pixels once.
- Always include the four neighbor pixels in case one is nearer.
- Fix projection matrix computation to align overscan pixels.
Pull Request: https://projects.blender.org/blender/blender/pulls/124735
All the arrays in the new Action DNA are named singularly (`layer_array`,
`fcurve_array` etc). `channelbags_array` was the only plural one. For
consistency, this is now also singular.
Note that this is a backward-incompatible change, and will effectively
erase all animation from layered Actions. No effort is taken to handle this
rename as the feature is still in its experimental phase.
Action and slot assignments are not affected by this change.
Apart from the lack of backward compatibility, no functional changes are
expected.
Pull Request: https://projects.blender.org/blender/blender/pulls/124768
Was using the old, not the modified name of the brush to build the brush
asset reference, that is used to identify the active brush asset.
The new brush was actually activated, but some of the UI like the asset
shelf would display it as such.
The fix contains two parts:
1. Grease Pencil v3 now stores stroke `init_time` as a float attribute,
it's not enough precisiton for unix timestamp. Now the time value is
truncated to allow better precision. It's still stored in seconds.
2. The previous logic for calculating stroke gap time of the build
modifier isn't correct, it used to only count two starting times as
gap time. Now it's fixed with correct delta time.
Pull Request: https://projects.blender.org/blender/blender/pulls/124350
