Use transform matrices instead of loc, rot, scale variables to store instance transforms.
Reviewed By: JacquesLucke
Differential Revision: http://developer.blender.org/D10211
This improves the fix introduced in rBdc8b31af578bfcf9f77dbce6d7b7da1006a3c8c1.
While it did not result in other bugs afaik, I thought it would be
good to make the fix more specific, in case some other code did
expect `layer->data` to be null. This updated fix checks for the
very specific case the fix was intended for.
The issue was that boolean custom data layers were not written to files,
because the dna struct name `bool` does not exist. Adding a struct that
just contains a `bool/uint8_t` does not seem to be possible, it looks like
the minimum dna struct size is 4 bytes.
The proposed solution has two parts:
1. Write the custom data layer using `BLO_write_raw` instead of
`BLO_write_struct_array_by_name`.
2. When loading a file, reinitialize any custom data layer that was
not saved correctly (this is just a fix for existing files).
Differential Revision: https://developer.blender.org/D10194
The `image_get_gpu_texture` didn't use the iuser->view_index but
recalculated the requested view again. This lead to inconsistent
behavior when switching between multi view textures or stereo textures.
This has been fixed by ensuring that the `iuser->view_index` is always
used.
An Image has only place to store 2 view textures. This is done for
right/left eye compositing. A multi view texture can have more views.
This would lead to reading and writing to unallocated space.
When a multiview texture is requested that is larger than 1. It will
always be cached as being the first eye. The `gpu_view` of the Image is
also used as a cache key to check this.
Dividing the workload by number of tasks in float is imprecise and
lead in some cases to particles not being calculated at all
(example: 20000 particles, 144 tasks).
Switching this calculation to integer makes sure we don't lose count.
Differential Revision: https://developer.blender.org/D10157
Things like pointers to particle systems, or softbody data being stored
outside of its modifier, make it impossible for internal modifier copy
data code to be self-contained currently. It requires extra processing.
In existing code this was handled in several different places, in
several ways, and alltogether fairly inconsistently. Some cases were
even not properly handled, causing e.g. crashes as in T82945.
This commit addresses those issues by:
* Adding comments about the hackish/unsafe parts `psys` implies when
copying some modifier data (since we need to ensure particle system
copying and remapping of those pointers separately).
* Adding as-best-as-possible handling of those cases to
`BKE_object_copy_modifier` (note that it remains fragile, but is
expected to behave 'good enough' in any practical usecase).
* Remove special handling for specific editor code
(`copy_or_reuse_particle_system`). This should never have been
accepted in ED code area, and is now handled by
`BKE_object_copy_modifier`.
* Factorize copying of the whole modifier stack into new
`BKE_object_modifier_stack_copy`, now used by both `object_copy_data`
and `BKE_object_link_modifiers`.
Note that this implies that `BKE_object_copy_modifier` and
`BKE_object_copy_gpencil_modifier` are now to be used exclusively to
copy single modifiers. Full modifier stack copy should always use
`BKE_object_modifier_stack_copy` instead.
Fix T82945: Crash when dragging modifiers in Outliner.
Maniphest Tasks: T82945
Differential Revision: https://developer.blender.org/D10148
`BKE_volume_is_loaded` uses `grids.filepath` to determine if the
grids are already loaded. The issue was that `grids.filepath` was
set before the grids were loaded, resulting in incorrect early
returns for other threads.
Differential Revision: https://developer.blender.org/D10150
Bevel caps always had incorrect normals, causing display glitches
in some cases.
It seems this never worked properly (at least 2.79 also had this bug).
Use the projection vector as the normal.
The double quaternion blending method in addition to the deformation
matrix of each bone requires their rest matrices. For ordinary bones
this literally should use the bone rest matrix without any ambiguity.
However, it was also using the bone rest matrix for all of its
B-Bone segments, which is incorrect and causes strange deformation
in some cases involving extreme non-uniform scale, especially
at boundaries between different B-Bones. There is also a similar
known issue that happens with scale at bending joints, and this
fix reduces the distortion when both bones are B-Bones.
This changes both the Armature modifier and the Armature constraint
to use the actual segment rest matrices. Unlike bones, these can have
scale even in rest pose, so normalization is required.
Differential Revision: https://developer.blender.org/D10003
The issue was that the mesh shared its vertex weights with the
original mesh (to reduce memory consumption). The solution is
to make a local copy of the vertex weights in this case.
Our beloved shapekeys are 'virtual' overrides, they need special
snowflake treatment here as well.
They do not have any override data, from override perspective they are
considered as mere sub-data from their owning ID (mesh, lattice, etc.).
Therefore, we should not copy override data from them, but instead
properly flag those new IDs as `LIB_EMBEDDED_DATA_LIB_OVERRIDE`.
Found while investigating T84373.
This commit adds the ability to provide a default value to
`attribute_try_get_for_output` and uses it for the `Point Scale` node,
which is important because the node uses multiplication.
The idea is to keep "name-specific" functionality in nodes rather than in
the attribute API, otherwise the complexity will be hard to keep track of.
So this fix doesn't apply to the Attribute Vector Math node, but hopfully
that is okay since that's now a lower level node for this purpose anyway.
Differential Revision: https://developer.blender.org/D10115
This fixes the behavior of some nodes when the same attribute
name is used for input and output. If both attributes have a
different type, they can't exist at the same time. Therefore,
the input attribute has to be removed in order to create the
output attribute.
Previously, the input attribute was remove before it was used
in any computations. Now, the output is written to a temporary
buffer and only later saved in the geometry component. This
allows both attributes to coexist within the node.
The temporary attribute is only create when necessary. The
normal case without name collisions still works the same
as before.
Differential Revision: https://developer.blender.org/D10109
Ref T83793.
The translate node moves every point in the geometry, and the scale
node multiplies the "scale" attribute of the input geometry by its input.
While these operations are already possible with the "Attribute" nodes,
these new nodes fit nicely with the nodes specifically for changing the
"rotation" attribute that already exist, and they provide a simpler way
to do the same thing.
Differential Revision: https://developer.blender.org/D10100
This change will try to add meta data when using a multilayered open
exr file output node in the compositor. It adds the current scene meta
data and converts existing cryptomatte keys so it follows the
naming that is configured in the file output node.
This change supports the basic use-case where the compositor is
used to output cryptomatte layers with a different naming scheme to
support external compositors. In this case the Multilayered OpenEXR
files are used and the meta data is read from the render result.
Meta data is found when render layer node is connected with the
file output node without any other nodes in between. Redirects and empty
node groups are allowed.
The patch has been verified to work with external compositors.
See https://devtalk.blender.org/t/making-sense-of-cryptomatte-usage-in-third-party-programs/16576/17
See patch for example files.
Reviewed By: Sergey Sharybin
Differential Revision: https://developer.blender.org/D10016
This adds a new Align Rotation to Vector node based on the mockup
in T83669.
Reviewers: HooglyBoogly, simonthommes
Differential Revision: https://developer.blender.org/D10081
Previously, the span returned by `WriteAttribute`s might not contain the
current value of the attribute for performance reasons. To avoid some
bugs, the span now always contains the old values (they might have to
be copied over from the internal storage, dependending on how the
attribute is stored).
The old behavior is still available with the `get_span_for_write_only`
method. The span that it returns might not contain the current
attribute values. Therefore, it should only be used when you want
to overwrite an attribute without looking at the old values.
When painting using Auto-Normalize or Lock Relative with some
groups locked, the locked weights may not add up precisely to
1 because of precision limitations, which results in creating
nonzero weights close to FLT_EPSILON. With Lock Relative display
mode this is very obvious and annoying (random red points amid
black or blue), so add an epsilon check to consider less than
1e-6 unlocked weight to be the same as 0.
In addition, in cases when no weight can be painted due to locks,
don't create vertex group entries at all if they don't exist yet.
Also, don't run Auto Normalize when not painting a deform group.
Differential Revision: https://developer.blender.org/D10000
* Simplify and clarify logic in `BKE_undosys_step_undo/redo_with_data_ex`,
by adding early return on invalid situations, renaming some variables,
and adding comments.
* Add more sanity checks in those functions.
No behavioral change are expected here, besides in potential edge-case,
invalid situations.
This is a preliminary change, before some deeper modifications of
`BKE_undosys` undo/redo API.
Differential Revision: https://developer.blender.org/D10033
This patch implements the same operations and interface as the regular
vector math node, but it runs for every element of the attribute. This
should expand what's possible with geometry nodes quite a bit.
Differential Revision: https://developer.blender.org/D9914
Root of the issue is that `BKE_object_eval_boundbox` (that ports back
evaluated bbox to orig object during active depsgraph evaluation) is
only called when object's geometry actually is evaluated.
However, with new undo, often Object itself can be changed by undo,
without requiring its geometry to be re-evaluated, which was leading to
the evaluated bbox not being copied back into orig object anymore.
Fixing that by moving bbox copying-to-orig code into
`BKE_object_sync_to_original` instead, which is always executed when
object is evaluated (as hinted by the comment above
`BKE_object_eval_boundbox` actually).
Also allows to cleanup code for armature eval, apparently.
Maniphest Tasks: T83422
Differential Revision: https://developer.blender.org/D9789
This can be useful as a workaround on the boundary with the pinned
vertices in some situations among other things, and completely copies
the existing design of the self collision vertex group setting.
Differential Revision: https://developer.blender.org/D10043
The `object.collision.use` flag was treated as a redundant marker
of the existence of the modifier, going as far as adding/removing
it when the value was changed, which is not actually very useful.
Removing the modifier loses its position in the stack, and requires
a dependency graph rebuild. It feels it may be a legacy flag?
What would be useful however is the ability to toggle collisions
dynamically without removing the modifier. This patch adjusts the
code to keep the modifier when the flag is disabled, and add it
if it doesn't exist when the flag is enabled. The modifier now
checks the flag at the start and quickly exits after cleaning
up stale data. The redesigned setting is exposed in the UI.
Collisions can't be disabled by simply using the modifier enable
flags because the modifier merely saves a snapshot of the mesh at
a certain point of the modifier stack for other objects to use,
and thus has to be able to clear the stale data.
Differential Revision: https://developer.blender.org/D10064
Previous code would not generate a session uuid for embedded IDs (like
root node trees or master collections). While this is not a problem
currently since those are not directly stored in Main data-base, this is
conceptually wrong, since those IDs still pertain the Main data.
Further more, this is blocking using `session_uuid` more in depsgraph in
place from ID pointer itself, as identifier (related to T84397).
No ID (even remotely) related to Main database should ever be created
directly through MEM_mallocN. Using `BKE_libblock_alloc` is the bare
minimum.
Note that there is no behavior change expected here.
Previously, it was only forbidden to delete the last scene. This can
lead to the situation where a .blend file only contains linked scenes.
This is problematic, because linked data might not always be available
or can be removed from a .blend file without having an additional check
for remaining scenes.
Now there always has to be at least one local scene.
Reviewers: mont29
Differential Revision: https://developer.blender.org/D10049
Currently such triangles are effectively already excluded, because
the calculated forces are not applied to pinned vertices. However
these forces are still being computed, which is inefficient.
This adds an early check for triangles where all vertices are
pinned during BVH overlap detection, which significantly speeds
up certain use cases with big fully pinned areas that happen to
overlap a collider. In case of self collision both triangles must
be fully pinned to exclude safely, because the computation is
symmetric and handles two triangles at the same time.
Differential Revision: https://developer.blender.org/D10041
Instances are created with an "index" parameter used for persistence over
time through animation. Currently the geometry nodes instancer passes
the index in the array for this value, but the arrays created by the
"Point Distribution" node aren't necessarily stable in this way when
the input mesh is deformed. In D9832 we already mostly solved this
problem with an `id` attribute. The solution here is to create instances
with this attribute as well.
It's important to note that deforming the instanced points *after*
distribution will usually be a better solution for this problem. This
solution is likely still important though.
Differential Revision: https://developer.blender.org/D10024
'Caused'/revealed by rBd29a720c45e5: Operators that fully re-create the
mesh would previously rely on `sculpt_update_object` called from update
code to get required sculpt-specific data layers re-added to the new
mesh.
Now instead put all code adding data to orig mesh for sculpt purpose
into a new util function (`BKE_sculpt_ensure_orig_mesh_data`), and call
that function when entering sculpt mode, and from voxel remesher code.
This is contonuing effort to more clearly separate orig data from evaluated
data handling/usage in sculpt code.
TODO: there are likely other code paths that would need to call that
new function?
Reviewers: @sergey, @pablodp606
Subscribers:
Rename it to mark it is only for `encode` callbacks, fix `encode`
callback of text undo to early fail in case it gets a NULL context, add
an assert to `BKE_undosys_step_push_with_type` that context is not NULL
when undotype requires a valid one.
Note that in practice this should not change anything, currently it
seems that we always get a valid context in
`BKE_undosys_step_push_with_type`?
This was caused when the BKE_pbvh_draw_cb function was used with
update_only_visible set to false. In that case, all nodes with the flag
were updating, but the update flag was only cleared for visible nodes.
This was causing constant updates per redraw in no visible nodes until
they enter the view frustum and their flag was cleared.
In order to fix this and prevent it from happening again:
- Updating the buffers, flushing the updates and clearing the flags are
now part of the same function. It does not make sense to do these in
separate places.
- The BKE_pbvh_draw_cb function was refactored so the
pbvh_update_draw_buffers is only called once. It should now be easier to
understand what the function does when it is used to update only visible
nodes or all nodes.
Reviewed By: mont29
Maniphest Tasks: T79146
Differential Revision: https://developer.blender.org/D9935
Stores cryptomatte hashes as meta data to the render result. Compositors could
use this for lookup on names in stead of hashes.
Differential Revision: https://developer.blender.org/D9553
Use the BMesh symmetrize operator instead of using the modifier code.
While we could support shape-keys with the existing code used by the
mirror modifier, we'd need to add code-paths for evaluated mesh & bmesh
conversion to handle shape-keys differently just for this one case,
since we want to avoid copying & processing shape-keys layers for
evaluated meshes in general.
