For some reason, the rework of liboverride handling of Collection items
insertion (rB33c5e7bcd5e5) completely missed to update accordingly the
default liboverride apply code...
Many thanks to Wayde Moss (@GuiltyGhost) for the investigation and
proposed solution.
This was caused by the use of a reserved keyword macro that is not
directly used but causes an error on some compiler.
Change the occurences to not match the macros.
This file was skipped by source/tools/utils/autopep8_clean.py
since it doesn't have a .py extension, running the autopep8 tool
recursively detects Python scripts without extensions.
- Increase the stack level so the reported line number references
script authors code (not Blender's wrapper function).
- Include the operator name and poll/call usage in the warning.
Support a way to temporarily override the context from Python.
- Added method `Context.temp_override` context manager.
- Special support for windowing variables "window", "area" and "region",
other context members such as "active_object".
- Nesting context overrides is supported.
- Previous windowing members are restored when the context exists unless
they have been removed.
- Overriding context members by passing a dictionary into operators in
`bpy.ops` has been deprecated and warns when used.
This allows the window in a newly loaded file to be used, see: T92464
Reviewed by: mont29
Ref D13126
These functions can be used with PyArg_ParseTupleAndKeywords
(and related functions) to take typed RNA arguments without
having to extract and type-check them separately.
No functional changes, extracted from D13126.
There's a small typo in the tool tip for applying the Parent Inverse. This patch fixes that typo
old:
{F13010751}
new:
{F13010749}
Reviewed By: Blendify
Maniphest Tasks: T97437
Differential Revision: https://developer.blender.org/D14693
This is mostly a cleanup to avoid hardcoding the eager calculation of
normals it isn't necessary, by reducing calls to `BKE_mesh_calc_normals`
and by removing calls to `BKE_mesh_normals_tag_dirty` when the mesh
is newly created and already has dirty normals anyway. This reduces
boilerplate code and makes the "dirty by default" state more clear.
Any regressions from this commit should be easy to fix, though the
lazy calculation is solid enough that none are expected.
This adds support for rendering motion blur for volumes, using their
velocity field. This works for fluid simulations and imported VDB
volumes. For the latter, the name of the velocity field can be set per
volume object, with automatic detection of velocity fields that are
split into 3 scalar grids.
A new parameter is also added to scale velocity for more artistic control.
Like for Alembic and USD caches, a parameter to set the unit of time in
which the velocity vectors are expressed is also added. For Blender gas
simulations, the velocity unit should always be in seconds, so this is
only exposed for volume objects which may come from external OpenVDB
files.
These parameters are available under the `Render` panels for the fluid
domain and the volume object data properties respectively.
Credits: kernel advection code from Tangent Animation's Blackbird based
on earlier work by Geraldine Chua
Differential Revision: https://developer.blender.org/D14629
The fix ensures that the reference count for `IShellItem *pSI` is decremented,
preventing a memory leak. For `IFileOperation *pfo` the decrement of the
reference count is only attempted when `CoCreateInstance` is successful.
Additionally, the gotos have been replaced with nested if/else statements.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D14681
Since shader sources are now parsed on demand via `GPUShaderCreateInfo`,
sources are not available to be read via
`GPU_shader_get_builtin_shader_code`.
Currently this results in a crash as the code tries to read `NULL`
pointers.
`GPU_shader_get_builtin_shader_code` was created with the intention of
informing the user how a builtin shader works, thus "replacing"
detailed documentation.
Therefore this function doesn't really have a practical use in an addon.
So, instead of updating the function (which would require several
changes to the gpu module), remove it and improve the documentation.
Release Notes: https://wiki.blender.org/wiki/Reference/Release_Notes/3.2/Python_API#Breaking_Changes
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D14678
Noted as part of T94775 investigation by Wayde Moss (@GuiltyGhost),
thanks!
NOTE: this mistake probably did not have any pratical impact in current
code, at least for overrides.
This is to make the codegen and shading nodes object type agnostic. This
is essential for flexibility of the engine to use the nodetree as it see
fits.
The essential volume attributes struct properties are moved to the
`GPUMaterialAttribute` which see its final input name set on creation.
The binding process is centralized into `draw_volume.cc` to avoid
duplicating the code between multiple engines. It mimics the hair attributes
process.
Volume object grid transforms and other per object uniforms are packed into
one UBO per object. The grid transform is now based on object which simplify
the matrix preparations.
This also gets rid of the double transforms and use object info orco factors
for volume objects.
Tagging @brecht because he did the initial implementation of Volume Grids.
Noticed while looking into oneAPI patch.
Seems to be unused, without clear indication why/when it might be
needed. Removing the function simplifies adding the new backend.
Differential Revision: https://developer.blender.org/D14652
Continuing the refactors described in T93602, this commit moves
the face dot tag set by the subdivision surface modifier out of
`MVert` to `MeshRuntime`. This clarifies its status as runtime data
and allows further refactoring of mesh positions in the future.
Before, `BKE_modifiers_uses_subsurf_facedots` was used to check
whether subsurf face dots should be drawn, but now we can just check
if the tags exist on the mesh. Modifiers that create new new geometry
or modify topology will already remove the array by clearing mesh
runtime data.
Differential Revision: https://developer.blender.org/D14680
The following methods weren't included in API docs.
- BlendDataLibraries.load
- BlendDataLibraries.write
- Text.region_as_string
- Text.region_from_string
This doesn't cause any functional change as the RNA property
of Bone wasn't overridden by the _GenericBone's property,
however the `children` property was documented twice, causing a warning.
The UI description for the `bpy.types.ActionFCurves.remove` was incorrect;
seemingly a copy-paste typo from the `rna_Action_groups_remove` function.
Reviewed By: sybren, Blendify
Differential Revision: https://developer.blender.org/D14659
Preserve multi socket link order when copying nodes or adding a new
group input sockets by linking directly to multi inputs from the group
input node's extension socket.
This is done by also copying the `multi_input_socket_index` when
the new links are created by copying existing or temporary links.
Reviewed By: Hans Goudey
Differential Revision: http://developer.blender.org/D14535
Keep the existing Rec.709 fit and convert to other colorspace if needed, it
seems accurate enough in practice, and keeps the same performance for the
default case.
Reimplement copy geometry node groups in C. The version implemented in
Python could also manually copy the animation data, but it's more
standard to do this with `BKE_id_copy_ex` and `LIB_ID_COPY_ACTIONS`.
Differential Revision: https://developer.blender.org/D14615
`rna_NodeSocket_refine` and `rna_Node_refine` take significant time
when building the `NodeTreeRef` acceleration data structure, but they
aren't used at all. This commit removes their eager calculation and
instead creates them on-demand in the `rna()` functions. They also
aren't inlined to avoid including `RNA_prototypes.h` in the header.
Differential Revision: https://developer.blender.org/D14674
Continued improvements to the new C++ based OBJ importer.
Performance: about 2x faster.
- Rungholt.obj (several meshes, 263MB file): Windows 12.7s -> 5.9s, Mac 7.7s -> 3.1s.
- Blender 3.0 splash (24k meshes, 2.4GB file): Windows 97.3s -> 53.6s, Mac 137.3s -> 80.0s.
- "Windows" is VS2022, AMD Ryzen 5950X (32 threads), "Mac" is Xcode/clang 13, M1Max (10 threads).
- Slightly reduced memory usage during import as well.
The performance gains are a combination of several things:
- Replacing `std::stof` / `std::stoi` with C++17 `from_chars`.
- Stop reading input file char-by-char using `std::getline`, and instead read in 64kb chunks, and parse from there (taking care of possibly handling lines split mid-way due to chunk boundaries).
- Removing abstractions for splitting a line by some char,
- Avoid tiny memory allocations: instead of storing a vector of polygon corners in each face, store all the corners in one big array, and per-face only store indices "where do corners start, and how many". Likewise, don't store full string names of material/group names for each face; only store indices into overall material/group names arrays.
- Stop always doing mesh validation, which is slow. Do it just like the Alembic importer does: only do validation if found some invalid faces during import, or if requested by the user via an import setting checkbox (which defaults to off).
- Stop doing "collection sync" for each object being added; instead do the collection sync right after creating all the objects.
Cleanup / Robustness:
This reworking of parser (see "removing abstractions" point above) means that all the functions that were in `parser_string_utils` file are gone, and replaced with different set of functions. However they are not OBJ specific, so as pointed out during review of the previous differential, they are now in `source/blender/io/common` library.
Added gtest coverage for said functions as well; something that was only indirectly covered by obj tests previously.
Rework of some bits of parsing made the parser actually better able to deal with invalid syntax. E.g. previously, if a face corner were a `/123` string, it would have incorrectly treated that as a vertex index (since it would get "hey that's one number" after splitting a string by a slash), instead of properly marking it as invalid syntax.
Added gtest coverage for .mtl parsing; something that was not covered by any tests at all previously.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14586
- Was not exporting "Poly" curves at all,
- Had a crash when a single object contains multiple curves of different types -- it had a check for "is this nurbs compatible?" only for the first curve, and then proceeded to treat the other curves as nurbs as well, without checking for validity.
Fixed both issues by doing the same logic as in the old python exporter:
- Poly curves are supported,
- Treat object as "nurbs compatible" only if all the curves within it are nurbs compatible.
Added test coverage in the gtest suite. While at it, made "all_curves" test use the "golden obj file template" style test, instead of a manually coded test that checks intermediate objects but does not check the final exported result.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14611
The new 3.1 OBJ exporter code had incorrect code to determine which vertex group a polygon belongs to -- for each vertex, it was only looking at the first vertex group it has, and not using the group weight either.
This 99% fixes T96824, but not 100% on the user's submitted mesh -- exactly two faces from that mesh get assigned a different group compared to the old exporter. Either choice is "correct" given that on these two faces there are two vertex groups with equal contribution. The old Python exporter was picking the group based on internal python group name map order, whereas the new C++ exporter is picking the group with the lowest index, in case of ties. I'm not sure if it's possible to fix this TBH, will have to wait until the importer is also C++.
While at it, the new vertex group calculation code was doing a lot of redundant work for each and every face (traversing group lists several times, allocating & freeing memory), so I fixed that. Exporting a 6-level subdivided Monkey mesh with 30 vertex groups was taking 810ms, now takes 330ms.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14500
This adds a basic unit test to check USD has been correctly
build with imaging components to support building both with
the old and new libs, it automatically adds the test when it
detects a library with imaging enabled. (platform devs will
have to pay attention it runs the test to validate the libs
build correctly)
For future use in the code it also defines a USD_HAS_IMAGING
define one could check if we're building against an USD lib
that has it (just because we build/ship with it, doesn't
mean downstream builds will ship with it, so we'll have
to be a little pro-active there)
Reviewed By: sybren
Differential Revision:https://developer.blender.org/D14456
In some circumstances singular files with numbers in their name (like
turntable-1080p.png or frame-1042.png) might be detected as a UDIM.
The root cause in this particular instance was because `BKE_image_get_tile_info`
believed this file to be a tiled texture and replaced the filename with
a tokenized version of it. However, later on, the code inside `image_open_single`
did not believe it was tiled because only 1 file was detected and our
tiled textures require at least 2. This discrepancy lead to the broken
filename situation.
This was a regression since rB180b66ae8a1f as that introduced the
tokenization changes.
Differential Revision: https://developer.blender.org/D14667
Also, as a suggestion, this patch changes Mask By Color and Color Filter to be the same shade of green as paint and smear tool icons
{F12998856}
{F12998857}
{F12998858}
Reviewed By: Julian Kaspar & Joseph Eagar
Differential Revision: https://developer.blender.org/D14632
Ref D14632
If the `wpoly` vector was small, the `wpoly_new` pointer could point
to part of its inline buffer on the stack, which becomes invalid out of
that scope. Instead, store `wpoly_new` as a span, and assign it properly
from the moved vector.
The HIG mentions that redundant words like "Enables" or "Activates"
shouldn't be used for tooltips of boolean properties. In this case
"When checked" was the redundant language that was implied by
the checkbox itself-- convention is to just state what the property
does when it's on.
Also change a few conjugations to the imperative and simplify
wording slightly, in order to be more consistent with language
elsewhere in Blender, and to be a bit more direct.
Differential Revision: https://developer.blender.org/D14644
Introduced by my recent commit: {rB3acbe2d1e933}
Lead to crash when insert_keyframe_direct() was called. Keyframing
crashed for NLA special properties (influence, animated_time),
driven properties, etc.
This commit changes the Curve to Mesh node to work with `Curves`
instead of `CurveEval`. The change ends up basically completely
rewriting the node, since the different attribute storage means that
the decisions made previously don't make much sense anymore.
The main loops are now "for each attribute: for each curve combination"
rather than the other way around, with the goal of taking advantage
of the locality of curve attributes. This improvement is quite
noticeable with many small curves; I measured a 4-5x improvement
(around 4-5s to <1s) when converting millions of curves to tens of
millions of faces. I didn't obverse any change in performance compared
to 3.1 with fewer curves though.
The changes also solve an algorithmic flaw where any interpolated
attributes would be evaluated for every curve combination instead
of just once per curve. This can be a large improvement when there
are many profile curves.
The code relies heavily on a function `foreach_curve_combination`
which calculates some basic information about each combination and
calls a templated function. I made assumptions about unnecessary reads
being removed by compiler optimizations. For further performance
improvements in the future that might be an area to investigate.
Another might be using a "for a group of curves: for each attribute:
for each curve" pattern to increase the locality of memory access.
Differential Revision: https://developer.blender.org/D14642
This is a partial fix to the fact that rendering with EEVEE or other GL
render engines is currently blocking the whole UI when asking to redraw
a viewport.
This patch just bypasses the viewport bind (containing the Draw Context
lock) and the following drawing. There is an update tagging to not
loose a viewport update if there was one asked.
Other queries other than view redraw (such as selection depth drawing or
offscreen drawing) will still block the whole UI as they need immediate
data feedback.
Ping @Severin for the change in `WM_draw_region_viewport_bind()`.
I'm assuming this is not an issue because it's highly unlikely to
bring up this operator during rendering. But in this case, it would just
lock as usual.
The bypassing in `DRW_notify_view_update` might be a bit overparanoid.
The ported normal calculation from ceed37fc5c neglected to
use the tilt attribute to rotate the normals around the tangents.
This commit adds that behavior back, adding a new math header file
to avoid duplicating the rotation function for normalized axes.
Differential Revision: https://developer.blender.org/D14655
This patch contains an initial pixel extractor for PBVH and an initial paint brush implementation.
PBVH is an accelleration structure blender uses internally to speed up 3d painting operations.
At this moment it is extensively used by sculpt, vertex painting and weight painting.
For the 3d texturing brush we will be using the PBVH for texture painting.
Currently PBVH is organized to work on geometry (vertices, polygons and triangles).
For texture painting this should be extended it to use pixels.
{F12995467}
Screen recording has been done on a Mac Mini with a 6 core 3.3 GHZ Intel processor.
# Scope
This patch only contains an extending uv seams to fix uv seams. This is not actually we want, but was easy to add
to make the brush usable.
Pixels are places in the PBVH_Leaf nodes. We want to introduce a special node for pixels, but that will be done
in a separate patch to keep the code review small. This reduces the painting performance when using
low and medium poly assets.
In workbench textures aren't forced to be shown. For now use Material/Rendered view.
# Rasterization process
The rasterization process will generate the pixel information for a leaf node. In the future those
leaf nodes will be split up into multiple leaf nodes to increase the performance when there
isn't enough geometry. For this patch this was left out of scope.
In order to do so every polygon should be uniquely assigned to a leaf node.
For each leaf node
for each polygon
If polygon not assigned
assign polygon to node.
Polygons are to complicated to be used directly we have to split the polygons into triangles.
For each leaf node
for each polygon
extract triangles from polygon.
The list of triangles can be stored inside the leaf node. The list of polygons aren't needed anymore.
Each triangle has:
poly_index.
vert_indices
delta barycentric coordinate between x steps.
Each triangle is rasterized in rows. Sequential pixels (in uv space) are stored in a single structure.
image position
barycentric coordinate of the first pixel
number of pixels
triangle index inside the leaf node.
During the performed experiments we used a fairly simple rasterization process by
finding the UV bounds of an triangle and calculate the barycentric coordinates per
pixel inside the bounds. Even for complex models and huge images this process is
normally finished within 0.5 second. It could be that we want to change this algorithm
to reduce hickups when nodes are initialized during a stroke.
Reviewed By: brecht
Maniphest Tasks: T96710
Differential Revision: https://developer.blender.org/D14504
* Curves objects now support the geometry nodes modifier.
* It's possible to use the curves object with the Object Info node.
* The spreadsheet shows the curve data.
The main thing holding this back currently is that the drawing code
for the curves object is very incomplete. E.g. it resamples the curves
always in the end, which is not expected for curves in general.
Differential Revision: https://developer.blender.org/D14277
**Relevant to Artists:** This patch adds an option to the Parenting
menu, `Object (Keep Transform Without Inverse)`, and Apply menu, `Parent
Inverse`. The operators preserve the child's world transform without
using the parent inverse matrix. Effectively, we set the child's origin
to the parent. When the child has an identity local transform, then the
child is world-space aligned with its parent (scale excluded).
**Technical:** In both cases, the hidden parent inverse matrix is
generally set to identity (cleared or "not used") as long as the parent
has no shear. If the parent has shear, then this matrix will not be
entirely cleared. It will contain shear to counter the parent's shear.
This is required, otherwise the object's local matrix cannot be properly
decomposed into location, rotation and scale, and thus cannot preserve
the world transform.
If the child's world transform has shear, then its world transform is
not preserved. This is currently not supported for consistency in the
handling of shear during the other parenting ops: Parent (Keep
Transform), Clear [Parent] and Keep Transform. If it should work, then
another patch should add the support for all of them.
Reviewed By: sybren, RiggingDojo
Differential Revision: https://developer.blender.org/D14581
The new created strokes were not setting the right `orig` pointers, so the select operator could not select the strokes and points.
Now, the original pointers are set in the new strokes. To set the values is necessary assign the original pointer from the reference stroke because in modifiers the stroke is evaluated, so we need back two levels: Eval->Eval->Orig
Issue introduced in {rB80859a6cb272}
Only seen on Windows.
The `keyword_parse` lambda function code did not consider `\r` as a
whitespace char, resulting in a wrong parse.
(More investigation needs to be done).
Assume that only one layer matches the id and return instead
of continuing to iterate over attributes after the layers have
been potentially reallocated.
This commit removes all EEVEE specific code from the `gpu_shader_material*.glsl`
files. It defines a clear interface to evaluate the closure nodes leaving
more flexibility to the render engine.
Some of the long standing workaround are fixed:
- bump mapping support is no longer duplicating a lot of node and is instead
compiled into a function call.
- bump rewiring to Normal socket is no longer needed as we now use a global
`g_data.N` for that.
Closure sampling with upstread weight eval is now supported if the engine needs
it.
This also makes all the material GLSL sources use `GPUSource` for better
debugging experience. The `GPUFunction` parsing now happens in `GPUSource`
creation.
The whole `GPUCodegen` now uses the `ShaderCreateInfo` and is object type
agnostic. Is has also been rewritten in C++.
This patch changes a view behavior for EEVEE:
- Mix shader node factor imput is now clamped.
- Tangent Vector displacement behavior is now matching cycles.
- The chosen BSDF used for SSR might change.
- Hair shading may have very small changes on very large hairs when using hair
polygon stripes.
- ShaderToRGB node will remove any SSR and SSS form a shader.
- SSS radius input now is no longer a scaling factor but defines an average
radius. The SSS kernel "shape" (radii) are still defined by the socket default
values.
Appart from the listed changes no other regressions are expected.
This adds a structure, `ABCReadParams`, to store some parameters passed
to `ABC_read_mesh` so we avoid passing too many parameters, and makes it
easier to add more parameters in the future without worrying about
argument order.
Differential Revision: https://developer.blender.org/D14484
Box, Circle and Lasso select were not taking into account if a
mask(point) was parented; selection was only succeeding in the original
place.
Now check coordinates from evaluated mask (points) instead while
setting selection flags and DEG tagging still happens on the original ID.
Maniphest Tasks: T97135
Differential Revision: https://developer.blender.org/D14651
Extremely subttle bug that would only appear in some specific
circumstances, would cause memfile undo writing code to falsely detect
some ID as changed because it would get the wrong 'starting point' of
comparison with existing previous memfile step.
See T85756 for detailed explanation and reproducible case.
This adds a new node editor overlay that helps users to see where
named attributes are used. This is important, because named
attributes can have name collisions between independent node
groups which can lead to hard to find issues.
Differential Revision: https://developer.blender.org/D14618
When GPU subdivision is enabled the mesh objects remain unsubdivided on
the CPU side, and subdivision should be requested somewhat manually (via
`BKE_object_get_evaluated_mesh`).
When referencing an object, the Object Info node calls
`bke::object_get_evaluated_geometry_set` which first checks if a Geometry
Set is present on the object (unless we have a mesh in edit mode). If so
it will return it, if not, the object type is discriminated, which will
return a properly subdivided mesh object via `add_final_mesh_as_geometry_component`.
The unsubdivided mesh is returned because apparently the check on the
Geometry Set always succeeds (as it is always set in `mesh_build_data`).
However, the mesh inside this Geometry Set is not subdivided.
This adds a check for a MeshComponent in this Geometry Set, and if one
exists, calls `add_final_mesh_as_geometry_component` which will ensure
that the mesh is subdivided on the CPU side as well.
Differential Revision: https://developer.blender.org/D14643
Originally was noticed when using a linked background scene and a scene
camera from another (local) scene.
The root issue was that relation from view layer to object's base flags
evaluation was using wrong view layer. This is because the relation was
created between object and currently built view layer, and it only was
happening once (since the object-level relations are only built once).
Depending on order in which `build_object` was called it was possible
that relation from a wrong view layer was used.
Now the code is better split to indicate which parts of object relations
are built when object comes from a base in the view layer, and which ones
are built on indirect linking of object to the dependency graph.
This patch makes relations correct in the cases when the same object is
used as a base in both active and set scenes. But, the operation which
handles object-level flags might not behave correctly as there is no
known design of what is the proper thing to do in this case. Making a
clear design and implementation of case when object is shared between
active and set scene is outside of the scope of this patch.
Differential Revision: https://developer.blender.org/D14626
Prefer using immVertex3f when 3D shaders are used for 2D rendering due to overhead of vertex padding in hardware. CPU overhead is negligible.
Authored by Apple: Michael Parkin-White
Ref T96261
Reviewed By: fclem
Maniphest Tasks: T96261
Differential Revision: https://developer.blender.org/D14494
MSL follows C++ standard convention, and such variable declarations within switch statements must have their scope localised within each case. Adding braces in all cases ensures correct behaviour and avoids 'case ... bypass initialization of local variable' compilation error.
struct initialisation to follow C++ rules for Metal. Implicit constructors replaced with either explicit constructors or list-initialization where appropriate.
Ref T96261
Authored by Apple: Michael Parkin-White
Reviewed By: fclem
Maniphest Tasks: T96261
Differential Revision: https://developer.blender.org/D14451
Add a new operator, "Start Tweaking Strip Actions (Full Stack)", which
allows you to insert keyframes and preserve the pose that you visually
keyed while upper strips are evaluating,
The old operator has been renamed from "Start Tweaking Strip Actions" to
"Start Tweaking Strip Actions (Lower Stack)" and remains the default for
the hotkey {key TAB}.
**Limitations, Keyframe Remapping Failure Cases**:
1. For *transitions* above the tweaked strip, keyframe remapping will
fail for channel values that are affected by the transition. A work
around is to tweak the active strip without evaluating the upper NLA
stack.
It's not supported because it's non-trivial and I couldn't figure it
out for all transition combinations of blend modes. In the future, it
would be nice if transitions (and metas) supported nested tracks
instead of using the left/right strips for the transitions. That
would allow the transitioned strips to overlap in time. It would also
allow N strips to be part of the (previously) left and right strips,
or perhaps even N strips being transitioned in sequence (similar to a
blend tree). Proper keyframe remapping through all that is currently
beyond my mathematical ability. And even if I could figure it out,
would it make sense to keyframe remap through a transition?
//This case is reported to the user for failed keyframe insertions.//
2. Full replace upper strip that contains the keyed channels.
//This case is reported to the user for failed keyframe insertions.//
3. When the same action clip occurs multiple times (colored Red to
denote it's a linked strip) and vertically overlaps the tweaked
strip, then the remapping will generally fail and is expected to
fail.
I don't plan on adding support for this case as it's also non-trivial
and (hopefully) not a common or expected use case so it shouldn't be
much of an issue to lack support here.
For anyone curious on the cases that would work, it works when the
linked strips aren't time-aligned and when we can insert a keyframe
into the tweaked strip without modifying the current frame output of
the other linked strips. Having all frames sampled and the strip
non-time aligned leads to a working case. But if all key handles are
AUTO, then it's likely to fail.
//This case is not reported to the user for failed keyframe
insertions.//
4. When using Quaternions and a small strip influence on the tweaked
Combine strip. This was an existing failure case before this patch
too but worth a mention in case it causes confusion. D10504 has an
example file with instructions.
//This case is not reported to the user for failed keyframe insertions. //
5. When an upper Replace strip with high influence and animator keys to
Quaternion Combine (Replace is fine). This case is similar to (4)
where Quaternion 180 degree rotation limitations prevent a solution.
//This case is not reported to the user for failed keyframe insertions.//
Reviewed By: sybren, RiggingDojo
Differential Revision: https://developer.blender.org/D10504
Undefined behaviour for divergent control-flow fixes, replacement for partial vector references, and resolution of a number of calculation precision issues occuring on macOS.
Authored by Apple: Michael Parkin-White
Ref: T96261
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D14437
This patch fixes a typo in commit e59f754c16 which incorrectly uses `GPU_TEXTURE_ARRAY` instead of `GPU_FORMAT_COMPRESSED`.
`GPU_FORMAT_COMPRESSED` and `GPU_TEXTURE_ARRAY` both currently evaluate to 16, so this patch does not change anything functionally; however, this patch will prevent issues from arising in the future.
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D14384
Add operator to select markers left/right of the current frame
(including the current frame).
`bpy.ops.marker.select_leftright(mode='LEFT', extend=False)`
`mode` can be either 'LEFT' or 'RIGHT'.
The naming and defaults of the above variables match similar operators
(e.g., `bpy.ops.nla.select_leftright`)
This also adds a new sub-menu to the Marker menu found in animation
editors, exposing both the new `bpy.ops.marker.select_leftright`
operator as well as the `bpy.ops.marker.select_all` operator.
Despite the name "Before Current Frame" and "After Current Frame", it
also selects a marker that falls on the current from for both of the
modes. This is to match the behavior found in the `nla.select_leftright`
operator.
RCS: https://blender.community/c/rightclickselect/OgmG/
Reviewed by: sybren, looch
Differential Revision: https://developer.blender.org/D14176
F2 allows renaming lots of different types of active items, and now it
also works for markers.
Before, Ctrl+M was used, and it's context-sensitive: you often get
"Mirror Keys" instead, when your cursor isn't on the markers region, and
that operator has nothing to do with either renaming or markers.
**What this commit does:**
- Replace Ctrl+M shortcut with F2.
- Adds the `TOPBAR_PT_name_marker` panel which is implemented similar
to the global rename panel. This having to press enter twice to
confirm or escape twice to cancel, which would happen if the
`marker.rename` operator was called directly.
- Replace usages of `marker.rename` in the UI with `wm.call_panel`.
- To make the Industry Compatible keymap consistent with Blender
Default, the rename shortcut only works when hovering the markers
area.
Reviewed By: ChrisLend, sybren
Differential Revision: https://developer.blender.org/D12298
There are no real difference from the previous icons, but since some
manual changes were introduced in the icons file, we still needed
a final sync between them.
Unlike regular selection cycling that is activated when clicking again
in the same location, object mode would cycle to another object
if the object that was selected happened to already be active.
This made it impossible to click-drag to tweak the active object
if there were other objects behind it as those would be activated first.
Resolves T96752.
Particles baked into memory would never load the final frame because
of an off-by-one error calculating the particles `dietime`.
This value indicates the frame which the particle ceases to exist but
was being set to the end-frame which caused this bug as the scenes
end-frame is inclusive.
While the last frame was properly written and read from memory,
the `dietime` was set to the last frame causing all the particles to be
considered dead when calculating the cached particle system.
The GPU evaluation for curves will have to change significantly from the
current particle hair drawing code, due to its more general use cases
and support for more curve types. To simplify that process and avoid
introducing regressions for the rendering of hair particle systems,
this commit splits drawing functions for the curves object and
particle hair.
The changes are just inlining of functions and copying code
where necessary.
Differential Revision: https://developer.blender.org/D14576
Problem is that the orco layer was not taken care of by the GPU
subdivision routines. This only handles the issues for EEVEE/Workbench.
For Cycles, this would need to be handled at the wrapper level somehow.
Use Alt-Slash to remove objects from local-view (was M prior to [0]),
following the convention of using Alt to perform the reverse of an
action. Also remove the confirmation menu as this key as it can be
undone and it's not likely to be pressed by accident.
This can be useful to quickly subtract items from a complex selection
with items that only become visible when entering local-view.
The M key was originally used in 2.4x since moving between layers wasn't
possible. Now moving between collections is possible in local-view
the keys collided.
[0]: cf5d582b77
edges
When wireframe mode is turned on, the subdivision edges not originating
from coarse edges were also drawn as regular edges, which would confuse
users trying to select them. These should not be drawn in edit mode,
only in object mode when optimal display is turned off (matching the CPU
subdivision case).
Fix word-wrapped tooltip text not showing by aligning to pixel grid.
See D14639 for more details.
Differential Revision: https://developer.blender.org/D14639
Reviewed by Campbell Barton
The "dir" argument to `BKE_where_on_path` was only actually
used in a few places. It's easier to see where those are if there
isn't always a dummy argument.
When displaying a deform modifier in edit mode, a cached
array of positions is used. Parallelizing bounds calculation when
that array exists can improve the framerate when editing slightly
(a few percent). I observed an improvement of the min/max itself
of about 10x (4-5ms to 0.4ms).
If all of the curves are poly curves, the evaluated positions are the
same as the original positions. In this case just reuse the original
positions span as the evaluated positions.
Addresses T97257, to make it consistent with regular attributes tab
Review by: Julian Kaspar
Differential Revision: https://developer.blender.org/D14631
Ref D14631
`GPU_shader_get_uniform_block` is marked as deprecated and the value
returned does not match what `GPU_uniformbuf_bind` expects.
Also, small typo fix in python error message.
Differential Revision: https://developer.blender.org/D14638
This behaviour was introduced in a687d98e67 to bring the old obscure
"M" operator to remove objects from the local view. In order to avoid
the keymap clash with the Move to Collection operator, the Move to
Collection was artificially restricted to work in local view.
In retrospect, the "Remove from Local View" operator is in the menu anyways,
so it didn't even need to have a shortcut (back in 2.79 the operator was
not in a menu).
The changes introduced here are:
* No shortcut for "Remove from Local View"
* No more restrictions to "Move/Link to Collection" from local view.
Thanks for Philipp Oeser for digging the old commit that introduced this
and for the rationale on the changes.
This can be useful to match transforms to what native Cycles
would see in Blender, as USD typically uses centimeters, but
Blender uses meters. This patch also fixes the hardcoded focal
length multiplicator, which is now using the same units as
everything else. Default of "stageMetersPerUnit" is 0.01 to match
the USD default of centimeters.
Differential Revision: https://developer.blender.org/D14630
Print it as a "%s" so that possible percentage symbols in the
error message does not cause issues.
Use proper assert (assert(true) is a no-op).
Also use `empty()` instead of `length()`.
Reviewed with Clement in real life.
Solves compilation warning with Clang, and moves manipulation with
DNA structures to the designed way for C++.
The tests and few other places are update to the new code by Jacques.
Ref T96847
Maniphest Tasks: T96847
Differential Revision: https://developer.blender.org/D14625
After appending, new link/append code would delete linked IDs, even if
those where pre-existing. Note that this would actually lead to invalid
memory access later in append code (ASAN crash).
This was one of multiple placeholder brushes to simplify development.
Having it is not necessary anymore.
It was a brush that could add new curves according to a specific density.
This functionality will be brought back as a new brush later.
Ref T97255.
Implements T97163
Newly created meshes have all voxel remesher checkboxes aside from Fix Poles enabled.
Startup files updated with versioning.
Reviewed By @JulianKaspar
Differential Revision: https://developer.blender.org/D14608
Ref D14608
Regression in [0] which caused canceled PRESS events not to generate
CLICK_DRAG.
Resolve by checking for an active brush tool in poll instead of the
PARTICLE_OT_brush_edit invoke function.
[0]: 4d0f846b93,
- Add logging for CLICK_DRAG event handling to debug drag events.
- Use logging API for reporting the key-map, operator and event.
This command now prints useful information for investigating
key-map and event handling issues:
blender --log "wm.handler.*" --log-level 4
Internally many offsets for BLF were integers but exposed as floats,
since these are used in pixel-space, many callers were converging them
back to integers. Simplify logic by using ints.
Support sub-pixel kerning and hinting for future support for improved
character placement. No user visible changes have been made.
- Calculate sub-pixel offsets, using integer maths.
- Use convenience functions to perform the conversions and hide the
underlying values.
- Use `ft_pix` type to distinguish values that use sub-pixel integer
values from freetype and values rounded to pixels.
This was originally based on D12999 by @harley with the user visible
changes removed so they can be applied separately.
In order to allow GLSL Cross Compilation across platforms, expose in
Python the `GPUShaderCreateInfo` strategy as detailed in
https://wiki.blender.org/wiki/EEVEE_%26_Viewport/GPU_Module/GLSL_Cross_Compilation
The new features can be listed as follows:
```
>>> gpu.types.GPUShaderCreateInfo.
define(
fragment_out(
fragment_source(
push_constant(
sampler(
typedef_source(
uniform_buf(
vertex_in(
vertex_out(
vertex_source(
>>> gpu.types.GPUStageInterfaceInfo.
flat(
name
no_perspective(
smooth(
>>> gpu.shader.create_from_info(
```
Reviewed By: fclem, campbellbarton
Differential Revision: https://developer.blender.org/D14497
Propagate the fp settings from the main thread to all the worker threads (the fp settings includes the FZ settings among other things) - this guarantees consistency in execution of floating point math regardless if its executed in tbb thread arena or on main thread
Add FZ mode to arm64/aarch64 in parallel to the way its been done on intel processors, currently compiling for arm target does not set this mode at all, hence potentially runs slower and with possible results mismatch with intel x86.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D14454
We better handle NULL object pointers before doing layer collections
resync, otherwise said resync process has to deal with those NULL
pointers. By the look of it this mistake has been there since the origin
of the remapping/relinking code.
Also for safety (and optimization), do not perform layer collection
resync from `libblock_remap_data_postprocess_object_update` when
`libblock_remap_data_postprocess_collection_update` is called
immediately afterwards.
Also added same 'skip on NULL collection object pointer' check to
`layer_collection_local_sync` as the one in
`layer_collection_objects_sync`, since it's fairly hard to always
guaranty there is no such NULL pointer when calling that code.
Animation would still play in the viewport.
There are two ways to unlink an action from the Outliner:
[1] `Unlink Action` on the Animation Data context menu.
This does `outliner_do_data_operation` / `unlinkact_animdata_fn` and has
the correct DEG update.
[2] `Unlink` on the Action context menu
This does `outliner_do_libdata_operation` / `unlink_action_fn` and was
missing the DEG update.
Now add the missing DEG update to the second case.
Maniphest Tasks: T95679
Differential Revision: https://developer.blender.org/D14089
The issue reported was that the recently introduced manual framerange of
an action (see rB5d59b38605d6) was not having an immediate effect in
case the action was pushed down from the Action Editor and only showed
its effects after e.g. saving and reloading the file. However doing the
same thing (pushing down the action) was working fine when done from the
NLA.
Now bring pushdown in sync (in terms of DEG update tagging) between the
Action Editor and the NLA, meaning that now both the owner and the
action are tagged when pushdown happens from the Action Editor as well.
Fixes T96964.
Maniphest Tasks: T96964
Differential Revision: https://developer.blender.org/D14564
The new tab in the properties panel for "Color Attributes" shows the
data type and domain, just like the Attributes tab.
The issue is that the UI does not scale well and can only display the
full names when dragged to an extreme width.
This is due to the inclusion of the render icon in between the
attribute name and attribute type description.
This patch changes the order that items are displayed in the Color
Attributes Panel. Moving the render icon to the very right.
The result is consistent with other parts of the Blender UI and does
not take as much space to display the full text.
Reviewed By: @jbakker
Differential Revision: https://developer.blender.org/D14567
Ref D14567
Color filter sharpening now clamps the output.
The sharpening delta is now calculated from the
difference of two levels of vertex averaging instead
of one smooth iteration and the base color.
TODO: Sharpen in a different color space;
SRGB-linear has saturation artifacts. I
tried HSL but it had value artifacts. I'd
like to try LAB but we don't seem to have
conversion functions for it (at least as far
as I could see).
- Replace SPACE_TYPE_LAST with SPACE_TYPE_NUM (adding 1).
- Rename RGN_TYPE_LEN to RGN_TYPE_NUM
This makes it possible to tag space-type/region-type combinations
with `bool tag[SPACE_TYPE_NUM][RGN_TYPE_NUM]` which reads more clearly
than `bool tag[SPACE_TYPE_LAST + 1][RGN_TYPE_LEN]`.
The type of the key was changed from string to integer in 66da2f537a, but
the GSet was still being created for string keys.
As long as the integers stay small enough, this even kind of works on
little-endian systems, but of course it should use an integer hash instead.
Currently, hovering over a socket itself shows no tooltip at all, while
hovering over its value field shows "Default value", which is not helpful.
This patch therefore implements socket tooltips following the proposal at
https://blender.community/c/rightclickselect/2Qgbbc/.
A lot of the basic functionality was already implemented for Geometry Nodes,
where hovering over the socket itself shows introspection info.
This patch extends this by:
- Supporting dynamic tooltips on labels, which is important for good tooltip
coverage in a socket's region of the node.
- Adding a function to setting a dynamic tooltip for an entire uiLayout, which
avoids needing to set it manually for a wide variety of socket types.
- Hiding the property label field in a tooltip when dynamic tooltip is also
provided. If really needed, this label can be restored through the dynamic
tooltip, but in all current cases the label is actually pointless anyways
since the dynamic tooltip gives more accurate and specific information.
- Adding dynamic tooltips to a socket's UI layout row if it has a description
configured, both in the Node Editor as well as in the Material Properties.
Note that the patch does not add any actual tooltip content yet, just the
infrastructure to show them. By default, sockets without a description still
show the old "Default value" tooltip.
For an example of how to add socket descriptions, check the Cylinder node
in the Geometry Nodes.
Differential Revision: https://developer.blender.org/D9967
When using proportional editing, the 'project onto self' snap setting
is ignored since proportional editing does not allow snapping to
self. The UI should reflect this fact. This patch makes 'project onto
self' active only when proportional editing is off.
Reviewed By: mano-wii
Differential Revision: https://developer.blender.org/D14496
Simply add the few lines so that the Cycles renderable New Curves
(with a small temporary patch to output the new type to render engines)
would get assigned a shader (in particular a hair shader).
Differential Revision: https://developer.blender.org/D14622
* Don't assume the display colorspace name fully defines the transform
to display space, this is not true in OpenColorIO 2 where view transforms
may be defined in more complexs ways than just specifying a colorspace.
* In places where we need to store the display colorspace name, resolve
<USE_DISPLAY_NAME> token manually.
Ref T96590
Adding nodes via `NodeAddOperator` could fail if the node's poll fails
in `rna_NodeTree_node_new`. Examples are trying to add a RenderLayers
node or a Cryptomatte node to a nodegroup.
Now except the python error and use blender error reporting only instead
of throwing full python stacktraces at the user.
Differential Revision: https://developer.blender.org/D14584
For some reasons(c) some base classes (like `bpy.types.Modifier`) are
not listed anymore by a call to
`bpy.types.ID.__base__.__subclasses__()`, unless they are first accessed
explicitely (e.g. by executing `bpy.types.Modifier` etc.).
Coordinate checks in `spline_under_mouse_get` need to take place with
the evaluated mask to get the right center. This is now more in line to
how this is done in `ED_mask_point_find_nearest`.
Note: similar issues are reported with box/circle/lasso selection in
T97135, will tackle these separately though.
Maniphest Tasks: T85467
Differential Revision: https://developer.blender.org/D14598
This implements the icon for snake hook, as well as tweak the
growth brush to make it bigger and with the triangles more
distinct for better reading of the icon.
Differential Revision: https://developer.blender.org/D14616
It was not possible to assign a shortcut to menu items in the insert
key-frame menu going back to version 2.7x. Doing so would replace the
current key that opens the insert keyframe menu (I-key by default),
instead of binding a key to insert a key-frame for the keying-set
referenced by the menu item.
Now each menu item can be bound to a key or added to the "Quick
Favorites" menu, directly inserting a key-frame for the corresponding
keying-set.
Note that users must use the operator `anim.keyframe_insert_by_name`
when setting up key-shortcuts as `anim.keyframe_insert` is only intended
to launch the menu.
Keymap Editor:
When editing these key-map items in the key-map editor, the keying-set
identifier must be used. At the moment the key-map editor doesn't
support showing a drop-down list. The identifiers can be used from the
tool-tip or the info editor.
{F12994924}
Details:
Use `ANIM_OT_keyframe_insert_by_name` instead of
`ANIM_OT_keyframe_insert_menu` for the insert keyframe popup menu to
resolve the following issues binding keys to keying sets:
- The index of the keying set isn't stable (adding/removing keying sets
may change it).
- Binding a key to items in the popup menu triggers a popup instead of
inserting a key using the keying set from the menu item.
While support for using the current operator could be improved, it will
still only work for built-in keying sets, so I'd prefer to use an
operator that is intended for key-bindings.
Besides supporting binding keys to menu items there are no functional
changes.
Reviewed By: sybren
Ref D14289
This simply adds access to the vertex crease layer from Python
code. Documentation was modified to remove "Edge" as it is
shared between edges and vertices, similarly to bevel weights.
Differential Revision: https://developer.blender.org/D14605
- Missing star prefix.
- Unnecessary indentation.
- Blank line after dot-points
(otherwise doxygen merges with the previous dot-point).
- Use back-slash for doxygen commands.
- Correct spelling.
These operators build a list of all lightgroups that are used by the view layer's
scene and either add all used lightgroups that are not part of the view layer yet
or remove all lightgroups in the view layer that are not being used.
Differential Revision: https://developer.blender.org/D14596
As far as I can see, it makes a lot of sense to have the alpha channel here, it matches the 2.x behavior and also matches what Eevee is doing.
Differential Revision: https://developer.blender.org/D14595
Use the Extend method for these, as these do not work correctly. For UVs
it's better to extend the UVs from the same face, and for tangent space
the normals should be encoded in a matching tangent space.
Later the Adjacent Faces method might be improved to support these cases.
Ref T96977
Differential Revision: https://developer.blender.org/D14572
Port the "Normal" and "Curve Tangent" nodes to the new curves data-block
to avoid the conversion to `CurveEval`. This should make them faster by
avoiding all that copying, but otherwise nothing else has changed.
This also includes a fix to move the normal mode as a built-in curve
attribute when converting to and from `CurveEval`. The attribute is
needed because the option is used implicitly in many nodes currently.
Differential Revision: https://developer.blender.org/D14609
Since effect strips can't be transformed directly, their selection had
to be forced in order to process them. This often failed in more
complicated scenarios, because there was no attempt to parse hierarchy
completely. In worst case only one effect in chain would be selected.
This code was marked by `XXX_DURIAN_ANIM_TX_HACK`.
Instead solution described above, a collection of strips that depend
on non effect strip position is built by function
`query_time_dependent_strips_strips` and it is stored in `TransSeq`.
In `flushTransSeq` this collection is iterated and transformation offset
is applied to effect strip animation. Strips in collection should be
consistent with true state of dependency and should be complete.
Functional changes:
- When 2-input effect strip changes position, animation is offset even
if only handles are moved. This only applies to 2-input effect however.
- Selection is not extended to include effect strips anymore. If effects
are to be moved, they must be selected manually. This is because
previously it was very hard to reorganize effects in chain, since moving
first strip in chain would always select anywhere from 1 to n effects.
So creating or filling gap in channel would almos always result in
collision especially if their order in timeline doesn't perfectly
represent their order in chain.
Previously, the new attributes were zero-initialized. However, sometimes
the default has to be something else. The old behavior led to unexpected
behavior in the Snap Curves to Surface operator in Deform mode, when the
curves were not attached to the surface before.
Differential Revision: https://developer.blender.org/D14588
This change moves the grid panel UI from the View tab up into the
Overlay panel.
Reasons to move to the Overlay panel include:
- Consistency with the grid options in the 3D viewport
- The grid has been drawn as an Overlay for quite some time already
Additional changes that now make sense to have:
- The grid responds to the main Overlay show/hide toggle
- Adds a toggle to show/hide the grid which is consistent with overlays in general
As before, these grid controls are only available for active UV edit
sessions.
Differential Revision: https://developer.blender.org/D11862
- Use "curve" instead of "spline" in comments
- Use non-plural variable names
- Tag topology dirty after resolution modified rather than positions
- Reorder enum values to change which value is zero (and the default)
- Remove a duplicate unused variable
draw_common.h was included in a C++ file
leading to the linker looking for the
decorated name for `G_draw` which lead
to a linker error.
adding an extern "C" for C++ fixes
the issue.
lengths along a set of points. This can be used for the sample curves
node, or finding new points along a curve when extending
or shrinking it.
This commit uses it in the snake hook brush as an example.
The logic is similar to the uniform length sampling, but the next
sample length is retrieved from the input instead of multiplication.
For the sample node in the future, though this sort of sampling can be
potentially done more efficiently for specific curve types besides
poly curves, it's simpler, at least as a start, to work on a set of
evaluated points that can be treated like a poly curve.
Differential Revision: https://developer.blender.org/D14571
Avoid installing all the ogg, theora, xvid etc. codec lib dev packages
unless we actually build ffmpeg itself. Otherwise they are not necessary
for Blender build itself.
This adds support to show dots for the curves points when in edit mode,
using a specific overlay.
This also adds `DRW_curves_batch_cache_create_requested` which for now
only creates the point buffer for the newly added `edit_points` batch.
In the future, this will also handle other edit mode overlays, and
probably also replace the current curves batch cache creation.
Maniphest Tasks: T95770
Differential Revision: https://developer.blender.org/D14262
The menu with the options was not visible because the tool checked must be the sculpt, not draw.
This was broken in old version, but I cannot determine when or if never worked at expected.
The `frame_offset` used for creating `TimeSamplings` when exporting was
being clamped, which would make subframe sampling potentially fail, or
get out of sync.
Both the Alembic and USD libraries use double precision floating
point numbers internally to store time. However the Alembic I/O
code defaulted to floats even though Blender's Scene FPS, which is
generally used for look ups, is stored using a double type. Such
downcasts could lead to imprecise lookups, and would cause
compilation warnings (at least on MSVC).
This modifies the Alembic exporter and importer to make use of
doubles for the current scene time, and only downcasting to float
at the very last steps (e.g. for vertex interpolation). For the
importer, doubles are also used for computing interpolation weights,
as it is based on a time offset.
Although the USD code already used doubles internally, floats were used
at the C API level. Those were replaced as well.
Differential Revision: https://developer.blender.org/D13855
This patch adds color attributes to TexPaintSlot. This allows an easier selection
when painting color attributes.
Previously when selecting a paint tool the user had to start a stroke, before the
UI reflected the correct TexPaintSlot. Now when switching the slot the active
tool is checked and immediate the UI is drawn correctly.
In the future the canvas selector will also be used to select an image or image texture node
to paint on. Basic implementation has already been done inside this patch.
A limitation of this patch is that is isn't possible anymore to rename images directly from
the selection panel. This is currently allowed in master. But as CustomDataLayers
aren't ID fields and not owned by the material supporting this wouldn't be easy.
{F12953989}
In the future we should update the create slot operator to also include color attributes.
Sources could also be extended to use other areas of the object that use image textures
(particles, geom nodes, etc... ).
Reviewed By: brecht
Maniphest Tasks: T96709
Differential Revision: https://developer.blender.org/D14455
Avoid Blender overwriting artist's choices. The automatic change from
"Hold" (i.e. bidirectional extrapolation) to "Hold Forward" (i.e. only
extrapolate forward in time) has been removed.
This patch does not change strip evaluation. Between two strips, the
first with `None` extrapolation and the next with `Hold`, neither strip
will evaluate, which matches previous behavior. A future patch can
change the evaluation behavior.
Reviewed By: RiggingDojo, sybren
Maniphest Tasks: T82230
Differential Revision: https://developer.blender.org/D14230
Previous commit (rBrB59681a7ccdcf) was effectively doing nothing, due to
weird hacks we had to do with OpenVDB 8.0 to 'integrate' NanoVDB.
Now OpenVDB 9.0 natively includes NanoVDB, which allows us to greatly
simplify that part of the code in install_deps.
The all_objects.blend test scene (in subversion tests repo) contained an
object with a subdivision surface. Which changes vertex positions
slightly, depending on used OpenSubDiv version and the compile flags. It
seems that the intent of the test was "test export of meshes that use
modifiers", so I changed that object to be a cube with a simple "taper"
modifier instead.
While at it, changed OBJ exporter test code to always print the
"expected and what we got" text difference details, when a test fails.
Much easier to see than just "the files are different" output. The code
to print that was behind an off by default flag for some reason.
This diff should get comitted together with updated all_objects templates
in subversion tests repo.
Reviewed By: Sebastian Parborg
Differential Revision: https://developer.blender.org/D14597
The problem was the original file had some vertex weight information, but the weights array was empty, so the duplication was not done and the free memory crashed.
To avoid this type of errors, now before duplicate weights the function checks the pointer and also the number of weights elements in the array to avoid the duplicatiopn of empty data.
With the increased use of multi-character format units and keyword-only
arguments these are increasingly difficult to make sense of.
Split the string onto multiple lines, one per argument.
While verbose it's easier to understand and add new arguments.
Currently, only Lightgroups that exist in the current view layer can be
selected from object or world properties.
The internal UI code already has support for search fields that accept
unknown input, so I just added that to the API and use it for lightgroups.
When a lightgroup is entered that does not exist in the current view layer
(e.g. because it's completely new, because the view layer was switched or
because it was deleted earlier), a new button next to it becomes active and
adds it to the view layer when pressed.
Differential Revision: https://developer.blender.org/D14540
* Add missing GLEW and hgiGL libraries for Hydra
* Fix wrong case sensitive include
* Fix link errors by adding external libs to static Hydra lib
* Work around weird Hydra link error with MAX_SAMPLES
* Use Embree by default for Hydra
* Sync external libs code with standalone
* Update version number to match Blender
* Remove unneeded CLEW/GLEW from test executable
None of this should affect Cycles in Blender.
Ref T96731
This seems to serve no purpose anymore, I don't see anywhere that
CD_MFACE is requested for modifier evaluation, and it's confusing
to have this in this final normals computation function.
Found while looking into D14579.
Differential Revision: https://developer.blender.org/D14580
The mechanism to instance meshes when there are no modifiers did not take
into account that modifiers might get re-evaluated from an operator that
requests loop normals. Now check for that case and no longer use the
instance then.
In the future, a better solution may be to compute loop normals on demand
as is already done for poly and vertex normals, but that would be a big
change.
Differential Revision: https://developer.blender.org/D14579
This frequently showed up in profiling but shouldn't.
This also updates the code to use atomics for more correctness and
adds multi-threading for better performance.
This implements two optimizations:
* Reduce virtual function call overhead when a non-standard virtual
array is used as input.
* Use a lambda in `type_conversion.cc`.
In my test setup, which creates a float attribute filled with the index,
the running time drops from `4.0 ms` to `2.0 ms`.
Differential Revision: https://developer.blender.org/D14585
I observed a 4-5x performance improvement (from 50ms to 12ms)
with five million points, though obviously the change depends on
the hardware.
In the future we may want to disable the parallelization in
`parallel_invoke` when there is a small amount of points.
Differential Revision: https://developer.blender.org/D14590
This patch adds an option to only use every n-th segment of the
envelope result. This can be used to reduce the complexity of the
result.
Differential Revision: http://developer.blender.org/D14503
Method which overrides a base class's virtual methods are expetced to
be marked with `override`. This also gives better idea to the developers
about what is going on.
The first element of the iterator was not being tested against the flag.
So in some cases it would lead to more objects been made into
single-user than the active (or selected) ones.
The goal is to make the Add menu more convenient for the new curves object.
The following changes are done:
* Add `curves` submenu.
* Add an `Empty Hair` operator that also sets the surface object.
* Rename the old operator to `Random`. It's mostly for testing at this point.
Differential Revision: https://developer.blender.org/D14556
This operator snaps the first point of every curve to the corresponding
surface object. The shape of individual curves or their orientation is
not changed.
There are two different attachment modes:
* `Nearest`: Move each curve so that the first point is on the closest
point on the surface. This should be used when the topology of the
surface mesh changed, but the shape generally stayed the same.
* `Deform`: Use the existing attachment information that is stored
for curves to move curves to their new location when the surface
mesh was deformed. This generally does not work when the
topology changed.
The purpose of the operator is to help setup the "ground truth"
for how curves are attached to the surface. When the ground
truth surface changed, the original curves have to be updated
as well. Deforming curves based on an animated surface will be
done with geometry nodes independent of the operator.
In the UI, the operator is currently exposed in curves sculpt mode
in the `Curves > Snap Curves to Surface` menu.
Differential Revision: https://developer.blender.org/D14515
Avoid errors in the legacy Pose Library panel (in Armature properties)
when the Pose Library add-on is disabled.
It's unfortunate that a built-in panel now has knowledge of an add-on.
Then again, it's temporary (one or two Blender releases), and it now uses
feature detection instead of just assuming the add-on is enabled.
The operator could crash in case the context "object" was overridden
from python, but the "active_object" wasnt (and the active object was
not a mesh).
Reason for the crash is a mismatch in the operators poll function
`data_transfer_poll` vs. `dt_layers_select_src_itemf` -- in the former,
the overriden "object" was respected (and if this was a mesh, the poll
was permissive), in the later it wasnt and only the "active_object" was
used (if this was not a mesh, a crash would happen trying to get an
evaluated mesh).
Now rectify how the object which is used is being fetched -> use
`ED_object_active_context` everywhere (see also rBe560bbe1d584).
Maniphest Tasks: T96888
Differential Revision: https://developer.blender.org/D14552
This does two things:
* Introduce new `materialize_compressed` methods. Those are used
when the dst array should not have any gaps.
* Add materialize methods in various classes where they were missing
(and therefore caused overhead, because slower fallbacks had to be used).
This improves performance e.g. when creating an integer attribute
based on an index field. For 4 million vertices, I measured a speedup
from 3.5 ms to 1.2 ms.
Add the ability to get/set the selected text.
**Calling the new methods:**
- `bpy.data.texts["Text"].region_as_string()`
- `bpy.data.texts["Text"].region_from_string("Replacement")`
Expose the "Connected" mode from the weld modifier in the
"Merge by Distance" geometry node. This method only merges
vertices along existing edges, but it can be much faster
because it doesn't have to build a KD Tree of all selected
points.
Differential Revision: https://developer.blender.org/D14321
Change the modifier name in the modifier stack to "Curvature 3D"
to be consistent with the modifier name in the drop-down.
Differential Revision: https://developer.blender.org/D14476
If all islands had a size of zero, a division by zero would occur in
`GEO_uv_parametrizer_pack`, causing the UV coordinates to be set to
NaN. An alternative approach would be to skip packing islands with a
zero size, but If UV coordinates are for example outside the 0-1 range,
it's better if they get moved into that range.
Differential Revision: https://developer.blender.org/D14522
Set the curve resolution to Bezier and Nurbs curves when converting
data using `curves_to_curve_eval`. This was missed in 9ec12c26f1.
Differential Revision: https://developer.blender.org/D14577
Add "for_write" on function names that retrieve mutable data arrays.
Though this makes function names longer, it's likely worth it because
it allows more easily using the const functions in a non-const context,
and reduces cases of mistakenly retrieving with edit access.
In the long term, this situation might change more if we implement
attributes storage that is accessible directly on `CurvesGeometry`
without duplicating the attribute API on geometry components,
which is currently the rough plan.
Differential Revision: https://developer.blender.org/D14562
Change uses of "Hair" in Render Settings UI in the property editor
and the "Hair Info" node to use the "Curves" name to reflect the
design described in T95355, where hair is just a use case of a more
general curves data type.
While these settings still affect the particle hair system,
the idea is that if we have to choose one naming scheme to align
with, we should choose the option that aligns with future plans
and current development efforts, especially since the particle
system is considered a legacy feature.
A few notes:
- "Principled Hair BSDF" is not affected since it's meant for hair.
- Python API property identifiers are not affected.
Differential Revision: https://developer.blender.org/D14573
The last length value was not initialized, and all length values were
moved one position towards the front of each curve incorrectly.
Also fix an assert when a curve only had a single point.
This revision allows to specify CUDA host compiler (nvcc's -ccbin command
line option) when configuring the build. It addresses the case where the
C/C++ compiler to be used in CUDA toolchain should be different from the
default C/C++ compiler, for instance in case of compilers versions conflicts
or multiple installed compilers.
The new CMake option is named `CUDA_HOST_COMPILER` and can be used as follows:
`cmake -DCUDA_HOST_COMPILER=<path-to-host-compiler>`
If the option is not specified, the build configuration behaves as previously.
Differential Revision: https://developer.blender.org/D14248
When showing an action data-block added to a library overridden object
in the Graph Editor, the visibility toggles would be disabled.
Toggling the visibility should be possible still and works with the
shortcuts, just the button was incorrectly disabled.
Also added the usual disabled hint for the tooltip.
Differential Revision: https://developer.blender.org/D14568
Reviewed by: Bastien Montagne
This commit disables the 'last updated' value (which is the date the
sphinx doc is generated), and instead modifies the 'commit' field from
the 'html_context' data to get:
- a link to the commit itself.
- the date of that commit.
This avoids having the whole documentation detected as changed every
time it is re-generated by the buildbot.
Reviewed By: dfelinto, campbellbarton
Differential Revision: https://developer.blender.org/D14429
Original rework of caches during undo/redo (see D8183) had a very bad
flaw hidden in it: using the key of a ghash as source of data.
While this was effectively working then (cache pointer itself being part
of the key, and said cache pointers not being cleared on file write),
this is a general very bad way to do things.
Now that cache pointers are more and more cleared on file write (as part
of clearing runtime-data to reduce false-positives when checking if an
ID has changed or not), this has to be fixed properly by:
* Not storing the cache pointer itself in the IDCacheKey.
* In undo context, in readfile code trying to preserve caches, store the
cache pointers as values of the mapping, together with the usages counter
The first change potentially affects all usages of
`BKE_idtype_id_foreach_cache`, but in practice this code is only used by
memfile reading code (i.e. undo) currently.
Related to T97015.
Reviewed By: brecht
Maniphest Tasks: T97015
Differential Revision: https://developer.blender.org/D14559
Python exceptions are now shown in the info editor,
this also resolves an old bug where errors were printed twice.
This was originally based on D9752 by @ShadowChaser although many
changes have been made from the original patch.
Details:
- BPy_errors_to_report no longer prints additional output.
- BKE_report_print_test was added so it's possible to check if calling
BKE_report also printed to the stdout.
- Callers to BPy_errors_to_report are responsible for ensuring output
is printed to the stdout/stderr.
- Python exceptions no longer add a trailing newline,
needed to avoid blank-space when displayed in the info-editor.
This avoids script authors using `eval("context.%s" % data_path)`
to access paths starting from the context,
which isn't good practice especially if the data_path isn't trusted.
Now it's possible to resplve paths such as:
context.path_resolve('active_object.modifiers[0].name')
Access the keys of the collection instead of the layers names
and use a set to detect collisions. There is no need to access the
duplicate layers themselves. Roughly twice as fast.
The `BVHCacheType bvh_cache_type` parameter defines specific
`BVHTrees` that cannot be customized.
So it doesn't make sense to pass this value to any
`*bvhtree_from_[...]_ex` function as the `BVHTrees` created in these
cases are custom and cannot be saved in the cache.
This also resulted in a nice cleanup in the code.
Differential Revision: https://developer.blender.org/D14479
In summary the changes are:
- Merge all `bvhtree_from_mesh_*.*_setup_data` in a single utility
- Create `bvhtree_from_editmesh_setup_data`
- Setup data only once in `BKE_bvhtree_from_mesh_get` and `BKE_bvhtree_from_editmesh_get`
Also the behavior of `BKE_bvhtree_from_mesh_get` and
`BKE_bvhtree_from_editmesh_get` changed a bit:
- If a null tree is cached, don't set the `data` to zero. This tree is not an error and the others data can still be used.
- If a null tree is returned, don't set the `data` to zero. Matches the previous change.
Differential Revision: https://developer.blender.org/D14549
color attribute system.
This commit removes sculpt colors from experimental
status and unifies it with vertex colors. It
introduces the concept of "color attributes", which
are any attributes that represents colors. Color
attributes can be represented with byte or floating-point
numbers and can be stored in either vertices or
face corners.
Color attributes share a common namespace
(so you can no longer have a floating-point
sculpt color attribute and a byte vertex color
attribute with the same name).
Note: this commit does not include vertex paint mode,
which is a separate patch, see:
https://developer.blender.org/D14179
Differential Revision: https://developer.blender.org/D12587
Ref D12587
This adds missing cases to detect edit mode for Curves objects.
Unlike other object types, Curves do not have specific edit data,
rather we edit the original data directly, and rely on `Object.mode`.
For this, `BKE_object_data_is_in_editmode` had to be modified to
take a pointer to the object. This affects two places: the outliner
and the dependency graph. For the former place, the object pointer
is readily available, and we can use it. For the latter, the object
pointer is not available, however since it is used to update edit
mode pointers, and since Curves do not have such data, we can
safely pass null to the function here.
This also fixes the assertion failure that happens when closing a file
in edit mode.
Differential Revision: https://developer.blender.org/D14330
Instead of using `CurveEval` to draw the curve wire edges, use
the new `Curves` data-block, which is already built as part of
an object's evaluated geometry set whenever there is a
`CurveComponent`.
This means that we can remove `Curve`'s temporary ownership
of `CurveEval` for drawing (added in 9ec12c26f1),
which caused a memory leak as described in T96498.
In my testing this improved performance by around 1.5x during
viewport playback, back to the performance of 3.1 before the
curve data structure transition started.
The next step of using the GPU to do the final curve evaluation
for the viewport is described in T96455, but is unrelated.
Differential Revision: https://developer.blender.org/D14551
This commit furthers some of the changes that were started in
rBb9febb54a492 and subsequent commits by changing the way surface
objects are presented to render engines and other users of evaluated
objects in the same way. Instead of presenting evaluated surface objects
as an `OB_SURF` object with an evaluated mesh, `OB_SURF` objects
can now have an evaluated geometry set, which uses the same system
as other object types to deal with multi-type evaluated data.
This clarification makes it more obvious that lots of code that dealt
with the `DispList` type isn't used. It wasn't before either, now it's
just *by design*. Over 1100 lines can be removed. The legacy curve
draw cache code is much simpler now too. The idea behind the further
removal of `DispList` is that it's better to focus optimization efforts
on a single mesh data structure.
One expected functional change is that the evaluated mesh from surface
objects can now be used in geometry nodes with the object info node.
Cycles and the OBJ IO tests had to be tweaked to avoid using evaluated
surface objects instead of the newly exposed mesh objects.
Differential Revision: https://developer.blender.org/D14550
Stumbled over the `integrate_surface_volume_only_bounce` kernel
function not returning the right type. The others too showed up as
warnings when building Cycles as a standalone which didn't have
those warnings disabled.
Differential Revision: https://developer.blender.org/D14558
Adds support for linking with some of the dependencies of a USD
build instead of the precompiled libraries from Blender, specifically
OpenSubdiv, OpenVDB and TBB. Other dependencies keep using the
precompiled libraries from Blender, since they are linked statically
anyway so it does't matter as much. Plus they have interdependencies
that are difficult to resolve when only using selected libraries from
the USD build and can't simply assume that USD was built with all
of them.
This patch also makes building the Hydra render delegate via the
standalone repository work and fixes various small issues I ran into
in general on Windows (e.g. the use of both fixed paths and
`find_package` did not seem to work correctly). Building both the
standalone Cycles application and the Hydra render delegate at the
same time is supported now as well (the paths in the USD plugin JSON
file are updated accordingly).
All that needs to be done now to build is to specify a `PXR_ROOT`
or `USD_ROOT` CMake variable pointing to the USD installation,
everything else is taken care of automatically (CMake targets are
loaded from the `pxrTargets.cmake` of USD and linked into the
render delegate and OpenSubdiv, OpenVDB and TBB are replaced
with those from USD when they exist).
Differential Revision: https://developer.blender.org/D14523
This adds a new Grow/Shrink brush which is similar to the Length
brush in the old hair system.
* It's possible to switch between growing and shrinking by hold
down ctrl and/or by changing the direction enum.
* 3d brush is supported.
* Different brush falloffs are supported.
* Supports scaling curves uniformly or shrinking/extrapolating
them. Extrapolation is linear only in this patch.
* A minimum length settings helps to avoid creating zero-sized curves.
Differential Revision: https://developer.blender.org/D14474
Those geometry types are expected to behave the same as e.g. mesh
with respect to data copying. The fact that this was not enabled
already was an oversight in the initial commit that added these types.
Differential Revision: https://developer.blender.org/D14554
Regression in [0] which missed excluding FRAME_CHANGE from
deg_recalc_flags_for_legacy_zero causing all
DEG_id_tag_update(&scene->id, 0) calls to re-calculate animation data.
When this tagging was done in the RNA update function, changing
key-framed values in the UI would be immediate reset to their
keyed-values.
Thanks to Philipp Oeser for finding the root cause.
[0]: 35aedd87e7
This is to improve grammatical consistency with other selection options.
Maniphest Tasks: T96745
Differential Revision: https://developer.blender.org/D14444
Support gizmos for the the action space type based on how it is done for
other types of spaces in Blender (e.g: view3d, image).
See patch submission for sample code.
Reviewed By: campbellbarton, sybren
Ref D13999
Some edit-curve operators used an 'ok' variable to represent
if the selection was found and if a change was made.
Previously it would only return cancel if an error was shown
causing a redundant undo step to be added without a selection.
Since this is simple behavior that shouldn't need much explanation,
use two variables with meaningful names to avoid confusion.
Reviewing D14511 highlighted this issue.
This experimental preference was not making much difference with
selection set to left-mouse for the 3D view and UV editor. Now dragging
the existing selection is possible when this preference is enabled.
Add a new panel called "Tweaks" so we can get feedback from users
about minor tweaks to behavior as exposing these minor changes.
Currently this only has a single item in it, however we may want to
get feedback from users about small changes in the future so I'd
prefer to have a place to list these kinds of options.
Ref D14542
Bug introduced in the smooth function changes done in commit rBd4e1458db3a0e0eaf80219dc8e6d10cb27620793
Differential Revision: https://developer.blender.org/D14548
The internal camera used to render the thumbnails also has to consider
`clip_start` and `clip_end`.
Reviewed By: Severin
Maniphest Tasks: T95678
Differential Revision: https://developer.blender.org/D14138
The legacy Pose Library operators now refer to the "Legacy Pose Library",
and their description mentions they are deprecated and will be removed
from Blender 3.3. The same was added to the `Object.pose_library` RNA
property.
Ref: T93405
The removal of these deprecated properties is tracked in T93406.
Remove much of the old legacy pose library:
- Remove from the Pose & Armature menus.
- Remove from the default keymap.
- Pose Library panel in Armature properties: Add "(legacy)" to title.
- Add note that the functionality of that panel is obsolete, with a
button that opens the manual on the chapter of the new pose library.
- Add button to convert the selected legacy pose library to pose assets.
- The rest of the functionality is greyed out to further communicate
it's been deprecated. It's still functional, though.
Ref: T93405
UV editor used wire color for drawing unselected vertices.
Add color variable to shader, so theme color can be used.
Reviewed By: jbakker
Differential Revision: https://developer.blender.org/D14373
Add edge panning feature to transform operator. It works in same way as
in node editor, but Y axis is limited by usable range up to 128
channels.
Reviewed By: Severin
Differential Revision: https://developer.blender.org/D14310
This only adds a experimental flag to enable the 3d texturing brush,
so future developments could check. Currently the flag does nothing
as no functionality of the 3d texturing brush has been implemented.
ImageTileWrapper is a wrapper around ImageTile to centralize tile calculations when
using CPP. Currentry used by the image engine and will be used for the 3d
texturing brush project.
This patch adds channel region to VSE timeline area for drawing channel
headers. It is synchronizedwith timeline region. 3 basic features are
implemented - channel visibility, locking and name.
Channel data is stored in `SeqTimelineChannel` which can be top-level
owned by `Editing`, or it is owned by meta strip to support nesting.
Strip properties are completely independent and channel properties are
applied on top of particular strip property, thus overriding it.
Implementation is separate from channel regions in other editors. This
is mainly because style and topology is quite different in VSE. But
also code seems to be much more readable this way.
Currently channels use functions similar to VSE timeline to draw
background to provide illusion of transparency, but only for background
and sfra/efra regions.
Great portion of this patch is change from using strip visibility and
lock status to include channel state - this is facilitated by functions
`SEQ_transform_is_locked` and `SEQ_render_is_muted`
Originally this included changes in D14263, but patch was split for
easier review.
Reviewed By: fsiddi, Severin
Differential Revision: https://developer.blender.org/D13836
This takes state of soc-2020-io-performance branch as it was at
e9bbfd0c8c (2021 Oct 31), merges latest master (2022 Apr 4),
adds a bunch of tests, and fixes a bunch of stuff found by said
tests. The fixes are detailed in the differential.
Timings on my machine (Windows, VS2022 release build, AMD Ryzen
5950X 32 threads):
- Rungholt minecraft level (269MB file, 1 mesh): 54.2s -> 14.2s
(memory usage: 7.0GB -> 1.9GB).
- Blender 3.0 splash scene: "I waited for 90 minutes and gave up"
-> 109s. Now, this time is not great, but at least 20% of the
time is spent assigning unique names for the imported objects
(the scene has 24 thousand objects). This is not specific to obj
importer, but rather a general issue across blender overall.
Test suite file updates done in Subversion tests repository.
Reviewed By: @howardt, @sybren
Differential Revision: https://developer.blender.org/D13958
This extracts the inner loops into a separate function.
There are two main reasons for this:
* Allows using `__restrict` to indicate that no other parameter
aliases with the output array. This allows for better optimization.
* Makes it easier to search for the generated assembly code,
especially with the `BLI_NOINLINE`.
Replace 5 arguments with a single struct as the same arguments
are used in many places.
This didn't read well and was confusing with both arguments named
`val` & `value` in the case of WM_modalkeymap_add_item.
Remove the extended version of ED_curve_editnurb_select_pick,
pass the size threshold directly to this function but as the distance in
pixels instead of a multiplier for ED_view3d_select_dist_px.
Using a multiplier is a less direct way to reference the threshold.
This has been broken for two years, since rB29f3af952725,
which retrieved the bounding box from an object and immediately
overwrote it with -1, 1. That commit had another problem though--
the modifier stack shouldn't use object level data, it should use
data from the previous modifier.
Differential Revision: https://developer.blender.org/D14524
Add support for the Curves object to the "Set Origin" and "Apply Object
Tansform" operators. Also change the automatic handle calculation to
avoid adding Bezier attributes if they don't need to be added.
Differential Revision: https://developer.blender.org/D14526
These functions are very simple, but some of them were showing up in
in profiles for curves sculpt mode and various curve nodes. Making sure
they are inlined will allow avoiding the compiler to optimize this logic
much better.
Differential Revision: https://developer.blender.org/D14529
This tool can be used to rapidly edit curves. The current set of
functionalities for Bezier splines are as follows:
The functionalities are divided into three versions of the operator:
* Left-Click
* Ctrl + Left-Click
* Double Click
All current functionalities and their defaults are as follows:
* Extrude Point: Add a point connected to an existing point.
Enabled for Left-Click.
* Extrude Handle Type: Type of the handles of the extruded points.
Can be either Vector or Auto. Defaults to Vector.
* Delete Point: Delete existing point.
Enabled for Ctrl + Left-Click.
* Insert Point: Insert a point into a curve segment.
Enabled for Ctrl + Left-Click.
* Move Segment: Move curve segment.
Enabled for Left-Click.
* Select Point: Select a single point or handle at a time.
Enabled for Left-Click.
* Move point: Move existing points or handles.
Enabled for Left-Click.
* Close Spline: Close spline by clicking the endpoints consecutively.
Defaults to True.
* Close Spline Method: The condition for Close Spline to activate.
Can be one of None, On Press or On Click.
Defaults to On Click for Left-Click and None for the others.
* None: Functionality is turned off.
* On Press: Activate on mouse down.
This makes it possible to move the handles by dragging immediately
after closing the spline.
* On Click: Activate on mouse release.
This makes it possible to avoid triggering the Close Spline
functionality by dragging afterward.
* Toggle Vector: Toggle handle between Vector and Auto handle types.
Enabled for Double Click on a handle.
* Cycle Handle Type: Cycle between all four handle types.
Enabled for Double Click on the middle point of a Bezier point.
The keybindings for the following functionalities can be adjusted from
the modal keymap
* Free-Align Toggle: Toggle between Free and Align handle types.
Defaults to Left Shift. Activated on hold.
* Move Adjacent Handle: Move the closer handle of the adjacent vertex.
Defaults to Left Ctrl. Activated on hold.
* Move Entire: Move the entire point by moving by grabbing on the handle
Defaults to Spacebar. Activated on hold.
* Link Handles: Mirror the movement of one handle onto the other.
Defaults to Right Ctrl. Activated on press.
* Lock Handle Angle: Move the handle along its current angle.
Defaults to Left Alt. Activated on hold.
All the above functionalities, except for Move Segment and
those that work with handles, work similarly in the case of Poly
and NURBS splines.
Reviewed By: HooglyBoogly, weasel, campbellbarton
Differential Revision: http://developer.blender.org/D12155
Moves all `interface_region*` files to C++ except for the tooptip region
which is slightly more complicated. Also move a few other files as well.
This helps to simplify and speed up code, especially through the use
of better C++ data structures. This change builds on all platforms on
the buildbot.
This patch re-adds the shading menu to lights to allow people to use lights in light groups.
This patch also hides all settings in the shading menu that are not useful for the light object.
Reviewed By: lukasstockner97
Maniphest Tasks: T96973
Differential Revision: https://developer.blender.org/D14527
Light groups are a type of pass that only contains lighting from a subset of light sources.
They are created in the View layer, and light sources (lamps, objects with emissive materials
and/or the environment) can be assigned to a group.
Currently, each light group ends up generating its own version of the Combined pass.
In the future, additional types of passes (e.g. shadowcatcher) might be getting their own
per-lightgroup versions.
The lightgroup creation and assignment is not Cycles-specific, so Eevee or external render
engines could make use of it in the future.
Note that Lightgroups are identified by their name - therefore, the name of the Lightgroup
in the View Layer and the name that's set in an object's settings must match for it to be
included.
Currently, changing a Lightgroup's name does not update objects - this is planned for the
future, along with other features such as denoising for light groups and viewing them in
preview renders.
Original patch by Alex Fuller (@mistaed), with some polishing by Lukas Stockner (@lukasstockner97).
Differential Revision: https://developer.blender.org/D12871
Also solves two warnings from the previous similar commit,
f688e3cc31. The change to the grease pencil
modifier is quite suspicious, but doesn't change the behavior,
which was already broken.
This adds support for selective rendering of caustics in shadows of refractive
objects. Example uses are rendering of underwater caustics and eye caustics.
This is based on "Manifold Next Event Estimation", a method developed for
production rendering. The idea is to selectively enable shadow caustics on a
few objects in the scene where they have a big visual impact, without impacting
render performance for the rest of the scene.
The Shadow Caustic option must be manually enabled on light, caustic receiver
and caster objects. For such light paths, the Filter Glossy option will be
ignored and replaced by sharp caustics.
Currently this method has a various limitations:
* Only caustics in shadows of refractive objects work, which means no caustics
from reflection or caustics that outside shadows. Only up to 4 refractive
caustic bounces are supported.
* Caustic caster objects should have smooth normals.
* Not currently support for Metal GPU rendering.
In the future this method may be extended for more general caustics.
TECHNICAL DETAILS
This code adds manifold next event estimation through refractive surface(s) as a
new sampling technique for direct lighting, i.e. finding the point on the
refractive surface(s) along the path to a light sample, which satisfies Fermat's
principle for a given microfacet normal and the path's end points. This
technique involves walking on the "specular manifold" using a pseudo newton
solver. Such a manifold is defined by the specular constraint matrix from the
manifold exploration framework [2]. For each refractive interface, this
constraint is defined by enforcing that the generalized half-vector projection
onto the interface local tangent plane is null. The newton solver guides the
walk by linearizing the manifold locally before reprojecting the linear solution
onto the refractive surface. See paper [1] for more details about the technique
itself and [3] for the half-vector light transport formulation, from which it is
derived.
[1] Manifold Next Event Estimation
Johannes Hanika, Marc Droske, and Luca Fascione. 2015.
Comput. Graph. Forum 34, 4 (July 2015), 87–97.
https://jo.dreggn.org/home/2015_mnee.pdf
[2] Manifold exploration: a Markov Chain Monte Carlo technique for rendering
scenes with difficult specular transport Wenzel Jakob and Steve Marschner.
2012. ACM Trans. Graph. 31, 4, Article 58 (July 2012), 13 pages.
https://www.cs.cornell.edu/projects/manifolds-sg12/
[3] The Natural-Constraint Representation of the Path Space for Efficient
Light Transport Simulation. Anton S. Kaplanyan, Johannes Hanika, and Carsten
Dachsbacher. 2014. ACM Trans. Graph. 33, 4, Article 102 (July 2014), 13 pages.
https://cg.ivd.kit.edu/english/HSLT.php
The code for this samping technique was inserted at the light sampling stage
(direct lighting). If the walk is successful, it turns off path regularization
using a specialized flag in the path state (PATH_MNEE_SUCCESS). This flag tells
the integrator not to blur the brdf roughness further down the path (in a child
ray created from BSDF sampling). In addition, using a cascading mechanism of
flag values, we cull connections to caustic lights for this and children rays,
which should be resolved through MNEE.
This mechanism also cancels the MIS bsdf counter part at the casutic receiver
depth, in essence leaving MNEE as the only sampling technique from receivers
through refractive casters to caustic lights. This choice might not be optimal
when the light gets large wrt to the receiver, though this is usually not when
you want to use MNEE.
This connection culling strategy removes a fair amount of fireflies, at the cost
of introducing a slight bias. Because of the selective nature of the culling
mechanism, reflective caustics still benefit from the native path
regularization, which further removes fireflies on other surfaces (bouncing
light off casters).
Differential Revision: https://developer.blender.org/D13533
With automatic collection previews (810e225c26) and a toggle for
collection instancing (previous commit) supported, there are no known
blocking issues for collection assets. There are still further
improvements to come as part of regular developemt (e.g. bounding box
based snapping).
Makes it possible to toggle instancing via the "Adjust Last Operation"
panel after dropping a collection asset into the viewport.
A design task that puts this into more context is pending still, but
this is a useful option to have either way.
Differential Revision: https://developer.blender.org/D14507
Reviewed by: Bastien Montagne
The Alembic procedural was only enabled during viewport renders
originally because it did not have any caching strategy. Now that
is does, we can allow its usage in final renders.
This also removes the `dag_eval_mode` argument passing to
`ModifierTypeInfo.dependsOnTime` which was originally added to detect if
we are doing a viewport render for enabling the procedural.
Differential Revision: https://developer.blender.org/D14520
This commit moves declarations that depend on `FN_field.hh` out of
`BKE_geometry_set.hh` into `BKE_geometry_fields.hh`. This helps to
reduce the number of areas that need to depend on the functions module,
which recently came in in review of D11591.
In the future we may have a library of standard field inputs in order to
make composing algorithms easier, so it makes sense to have a header
that could contain them and some basic related utilities relating the
concepts of geometry and fields.
Reducing use of unnecessary headers may also reduce compilation time.
Differential Revision: https://developer.blender.org/D14517
This commit adds attribute search the the attribute input field node.
Because it's a field node, finding which attribute to display without
increasing the complexity a lot isn't obvious. In this commit, all
attributes used by nodes in the current group are included.
When an attribute is chosen from the list, the node's data type is
updated, and links connected to the output socket are reconnected.
Ref T96271
Differential Revision: https://developer.blender.org/D14516
This commit ports the "Set Handle Positions" and "Set Hanle Type"
nodes to use the new curves data-block. The nodes become simpler
and likely much faster too, though they're usually not the bottleneck
anyway.
Most of the code is ported from `BezierSpline` directly. The majority
of the complexity comes from the interaction between different
automatically calculated handle types. In comparison `BezierSpline`,
the calculation of auto handles is done eagerly-- mostly because it's
simpler. Eventually lazy calculation might be good to add.
Differential Revision: https://developer.blender.org/D14464
Original report (T96763) only reported the issue of double-space before the texture path, but while adding test coverage I found some other issues that I fixed while at it:
- Incorrectly emits two spaces between `map_Xx` keyword and the texture path, leading to some 3rd party software not finding the textures,
- Emissive texture map (`map_Ke`) was not exported,
- When Mapping node is used on the texture UVs, the "Location" and "Scale" values were mixed up (location written as "scale", scale written as "location).
Added gtest coverage.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14519
PointCache handing is just horrible from RNA, makes dealing with
overrides a nightmare...
Ended up having to add a specific 'apply' callback for the `use_disk_cache`
property, that would explicitely NOT call the the `update` callback of
this property, to avoid having the whole disk cache nuked away...
But the whole thing remains fairly britle, to say the least.
While this is the desired behavior in almost cases, there are a few
hairy nightmares that may require not to do so.
NOTE: this change should should not modify any current behavior at all.
Not really clear why that would only show with multiple caches... But
point cache system is beyond brittle anyway.
This fix solves the issue at two different levels:
* General safety check in `rna_Cache_info_length` that we do get a valid
`pid`.
* Forbid usage of this `PointCache.info` RNA property in any
diffing/LibOverride processing (since it's by definition runtime, volatile
data).
Keep the value slider from clipping through rounded corners for
low values by ensuring the width of the slider rectangle is at least
twice the corner radius.
Reviewed By: Hans Goudey
Differential Revision: https://developer.blender.org/D11474
In Alembic curve topology is stored with an array of values describing
how many points each sub-curve has. Instead of writing the number of
points for the current curve, the Alembic exporter would write the
accumulated number of points.
This error has existed since the initial implementation.
This function was not used for anything other than mat4. This
was because of a limitation of the DRW module/
This makes it cleaner for the GLSL and also less tempting to use
it for other unconventional purpose.
This was caused by the recent changes made to the way we handle matrix
copies. The matrix copy assumed that the uniform iteration was the same
as creation order. But this was far from true. The reality was that
the iterator was reverse for `unichunk` but not for `unichunk->uniforms`
so this was recreating wrong matrix.
I rewrote this part to always use reverse iteration and fix the
copy destination.
Also I simplified the code making the assumption this won't be used for
anything else than mat4.
Adds a dropdown for the Library Overrides display mode that lets users
choose between a "Properties" and a "Hierachies" view mode. The former
is what was previously there (a mode that displays all overridden
properties with buttons to edit the values), the latter is new. It
displays the hierarchical relationships between library overridden
data-blocks. E.g. to override the mesh of an object inside a linked
collection, the entire collection > object > mesh hierarchy needs to be
overridden (whereby the former two will be automatically overridden
using system overrides).
The Hierarchies mode will also show the override hierarchies of
data-blocks that were linked and are overridden in the source file. This
information is useful to have, especially for debugging scenes.
Part of T95802.
Differential Revision: https://developer.blender.org/D14440
Reviewed by: Bastien Montagne
Direct replacement of code:
- memcpy of a single point is done as a shallow_copy() assignment.
- memcpy of a range of points is done with an explicit cast to void*
to tell compiler that we really want to memcpy even a non-trivial
type.
In some cases it seems that memcpy can be used more (points are copied
in a loop). Those left as-is since this is supposed to be a simple
cleanup.
Differential Revision: https://developer.blender.org/D14505
It's better to use some local/stable identifiier to avoid relying on
the data not being freed in between creating the search menu and
the exec function. This is similar to c473b2ce8b.
029cf23d71 changed some icons alignment, but after 9be49a1069
the icons don't align anymore. This commit reverts 029cf23d71 and
also makes a couple of other outliner icons left aligned, instead of
right aligned, for consistency and general alignment.
Differential Revision: https://developer.blender.org/D14501
Add a check to the creation of node groups to remove hidden links
that are connected to the outside of the node group. This avoids
creating sockets in the group's interface that aren't (visibly)
connected to anything within the node group.
Reviewed By: Jacques Lucke, Hans Goudey
Differential Revision: https://developer.blender.org/D14249
This commit makes the dot grid used as background in the node editor
more visually stable when zooming in and out.
The dot grid now uses a continuously subdividing pattern, where
each level of subdivision divides the previous five times, similar to
the line grid in the 3D viewport.
The maximum for the "Grid Levels" theme setting is changed to 3, since
any further subdivisions are too small to be visible.
The "Grid Levels" value for the default themes "Blender Dark" and
"Blender Light" is therefore changed to 3, as well.
Reviewed By: Hans Goudey, Pablo Vazquez
Differential Revision: http://developer.blender.org/D13302
- Metal uniform array compatibility in DRW module.
- Guard OpenGL-specific workarounds and flushes behind GPU_type_matches_ex API guard. Add further render boundaries for render paths called outside of the main loop.
Authored by Apple: Michael Parkin-White
Ref: T96261
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D14438
Skip validation when inserting items into the Win32 "Volumes" list.
This fixes some long hangs when launching Blender with disconnected
network shares.
See D14506 for more details.
Differential Revision: https://developer.blender.org/D14506
Reviewed by Brecht Van Lommel
MSL does not have an implicit global scope, this is emulated via macro's adding an indirection for uniforms, attributes, shader stage inputs and outputs such as:
#define roughness shaderinst->roughness.
Variables in GLSL which exist within uniform blocks can be directly referenced via the global scope, unlike standard C++. This means that variable name pollution occurs if subsequent local variables in the code use the same name, resulting in compilation errors.
A number of these conflicting names have been renamed to ensure unique naming and no further scope pollution.
Ref T96261
Reviewed By: fclem
Maniphest Tasks: T96261
Differential Revision: https://developer.blender.org/D14452
Explicit constructor for mat3 from a mat4 is not valid and cannot be overloaded.
Adding explicit texture resource type flags for depth textures. This is an explicit requirement for Metal Shading language. This is a temporary compatibility, as this path is already supported in GPU_SHADER_CREATE_INFO under ImageType::DEPTH_2D, though required in shader source for MSL shaders which do not have create info.
Authored by Apple: Michael Parkin-White
Ref T96261
Reviewed By: fclem
Maniphest Tasks: T96261
Differential Revision: https://developer.blender.org/D14418
This commit re-implements the resample curve node to use the new curves
type instead of CurveEval. The largest changes come from the need to
keep track of offsets into the point attribute arrays, and the fact
that the attributes for all curves are stored in a flat array.
Another difference is that a bit more of the logic is handled by
building of the field network inputs. The idea is to let the field
evaluator handle potential optimizations while making the rest of the
code simpler.
When resampling 1 million small poly curves,the node is about 6
times faster compared to 3.1 on my hardware (500ms to 80ms).
This also adds support for Catmull Rom curve inputs.
Differential Revision: https://developer.blender.org/D14435
The crash happens because the origindex layers created as part of
the modifier stack evaluation are not set in the `MeshRenderData` when
they should have been.
This is because when selecting in X-ray mode, a subdivision wrapper
is created to ensure that selection happens with a subdivided
geometry, and this replaces the `MDATA` wrapper which is also used to
setup the `MeshRenderData`.
As we do not seemingly have an `MDATA` wrapper, the draw code decides
that we can extract draw buffers directly from the BMesh, instead of
the mapped Mesh with origin indices layers.
To fix this, we should also consider to use mapped extraction if a
subdivision wrapper exists on the mesh.
Differential Revision: https://developer.blender.org/D14485
This is because the inheritance is not done before checking if the shader
should be statically compiled. Also some inheritance scheme
might have intermediate permutation that are not compilable.
The recent change to allow a max segments of 1000 in the modifier
causes a lag when dragging or wheeling in the segments box.
This change makes the soft limit back to 100, but you can still
type numbers up to 1000 in the box.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14477
There were multiple issues at hand here:
- The default value has been changed to `{0, 0, 1}` see: rB25f1783673de636a6f0ca4457df8c05bc685981a
- The output needs the subtype set `PROP_DIRECTION`
- The noder properties were missing in `node_composit_set_butfunc`
Fixes T96860
This adds a new operator that converts all selected curves objects
into hair particle systems on their respective surface objects. Existing
particle systems with the correct name are updated, otherwise a new
particle system is added.
The purpose of the operator is the make the new curve sculpting tools
useful even before all functionality is ported over from the old hair system.
The operator can be found in the `Object > Convert` menu in object mode,
when a curves object is active.
Differential Revision: https://developer.blender.org/D14441
When a `GPencilUpdateCacheNode` is created, it always allocates the
`children` pointer. This should not be freed until the whole cache is
deleted.
The `cache_node_update` would free the `children` pointer in a specific
case, causing a double-free later when the cache was removed.
The current behaviour is to prevent multi-user data from having its
transformation applied.
However in some particular cases it is possible to apply them:
* If all the users of the multi-user data are part of the selection.
* If not all the users are in the selection but the selection is made
single-user.
The active object is used as reference to set the transformation of the
other selected objects.
Note: For simplicity sake, this new behaviour is only available if all
the selection is using the same data.
Differential Revision: https://developer.blender.org/D14377
The current behaviour is to prevent multi-user data from having its
modifier applied.
Instead, with this patch, we now warn the user that if they want to
proceed the object will be made single-user.
Note that this only makes the object data single-user. Not the material
or actions.
As a future step we can apply the same behaviour for the Grease Pencil modifiers
Differential Revision: https://developer.blender.org/D14381
While it's useful for click-drag to leave the selection as-is
(when clicking on items that are already selected), it's useful
for a single click to de-select all other elements.
This also removes the need for the initial selection to set the object
as active since this is possible by clicking on it.
Using the evaluated lengths cache from 72d25fa41d, re-implement
the curve parameter node with the new data structure. Conceptually
it works the same way, but the code is restructured and cleaned up
a bit as well. This also adds support for Catmull Rom curves.
Differential Revision: https://developer.blender.org/D14461
The idea that curves with two points cannot be cyclic came from some
existing code, but there's not fundamental reason for it, so remove the
check in this function. The case can be handled elsewhere if necessary.
This commit adds calculation of lengths along the curve for each
evaluated point. This is used for sampling, resampling, the "curve
parameter" node, and potentially more places in the future.
This commit also includes a utility for calculation of uniform samples
in blenlib. It can find evenlyspaced samples along a sequence of points
and use linear interpolation to move data from those points to the
samples. Making the utility more general aligns better with the more
functional approach of the new curves code and makes the behavior
available elsewhere.
A "color math" header is added to allow very basic interpolation
between two colors in the `blender::math` namespace.
Differential Revision: https://developer.blender.org/D14382
Fix small cosmetic issues with the reroute node:
1. Remove special case that allowed curved links to attach vertically.
2. Center align the reroute node's label.
The vertically attached node links could lead to kinks in the otherwise
smooth curves. This would break the visual flow and make the link
potentially intersect the node's label.
The center alignment of the label gives more consistent results for
different label lengths and also reduces the chance of the label
interfering with the node links.
Reviewed By: Hans Goudey, Pablo Vazquez
Differential Revision: D14457
Don't always create a new geometry nodes node tree when adding a
geometry nodes modifier.
This avoids files getting cluttered with empty and unused geometry node
trees that are created every time a nodes modifier is added to an
object - even if only to apply an already existing.
This is also more consistent with other modifiers that also don't
automatically create new data blocks.
The new modifier still automatically gets populated with a new node
tree when adding it via the "New" button in the header of the
geometry nodes editor.
Reviewed By: Hans Goudey, Dalai Felinto, Pablo Vazquez
Differential Revision: D14458
Add new `BKE_id_is_editable` helper in `BKE_lib_id.h`, that supercedes
previous check (simple `ID_IS_LINKED()` macro) for many editing cases.
This allows to also take into account 'system override' (aka
non-editable override) case.
Ref: {T95707}.
'Delete' was a confusing name, even though it would delete the overrides
it would replace them by linked data.
Adding the 'single' version of that operation made it even more
confusing, since often it has to keep the override ID for sakes of
hierarchy, and just reset it and turn it back into a non-editable system
override.
Ref: {T95707}.
Implement default behavior to decide which overrides remain 'system'
ones, and which become 'user editable' ones, when creating hierarchy
override from 3DView or the Outliner.
3DView:
If from an Empty-instanced collection, only Armature objects in
that collection are user overrides.
If from a set of selected objects, all overrides created from selected
objects are user overrides.
Outliner:
All override IDs created from selected elements in the Outliner are user
overrides.
There is one special case: When a collection is selected, and is
'closed' in the outliner, all its inner armature objects are also user
overrides.
Ref: {T95707}.
When creating with hierarchies, core code only generates system
overrides, responsibility to define 'user overrides' is then for the
higher-level calling code (Editor/Operator-level).
do_version code uses fairly basic euristics, should be good enough here
though in most cases. and can always be refined later if needed.
Ref: {T95707}.
The whole liboverride data is still ignored by override diffing etc.,
but some of their flags should be editable (from script and/or advanced
technical/debug UI). So using a weird combination of flags to achieve
this.
Ref: {T95707}.
This merely adds the flag, exposes it in RMA, and uses it in some of the
most common 'is editable' checks (RNA, `BASE_EDITABLE` macro...).
Next step: do_version and defining systemoverrides at creation.
Ref: {T95707}.
The "dupli" system now has a faster, more powerful, and more flexible
alternative with geometry nodes. Since the point cloud objects haven't
been exposed in the non-experimental UI yet, we can remove the dupli
implementation and the panel for the object type.
Differential Revision: https://developer.blender.org/D14482
Adds fading support for build modifier so it's not a hard cut off
Reviewed By: Antonio Vazquez (antoniov), Matias Mendiola (mendio)
Differential Revision: https://developer.blender.org/D14309
The check for the selected status was missing in the case
where the stroke one has one point.
Differential Revision: http://developer.blender.org/D14490
Adds supports for collection previews that are rendered automatically when
collections are marked as assets. (Or when preview rendering is triggered
differently, e.g. through the //Refresh Data-Block Previews// operator).
Idea in this patch is to create a collection instance empty outside of main for
the collection, and then reuse the object rendering code to render the preview.
This keeps things very simple and works just fine.
Differential Revision: https://developer.blender.org/D14460
Reviewed by: Bastien Montagne
Error exposed by ba49345705. Code just assumed that the tree-element
pointed to a real ID, but this is often not the case, and the ID pointer
contains completely different data. E.g. before ba49345705, it would
be a pointer to one of the `Main` listbases, so this code would have
undefined behavior. Now the pointer is null for elements in the "Current
File" element, causing a null-pointer dereference rather than undefined
behavior (that just happened to virtually always result in the intended
code path).
The indirect library data icon was just a grayed out version of the
regular one. This graying out is now done in code, so the icon can be
removed from the SVG. Note that the icon is still defined as
`ICON_LIBRARY_DATA_INDIRECT` (or `LIBRARY_DATA_INDIRECT` in BPY).
The lines paint mask IBO extraction was not implemented for GPU subdivision.
For it to work, we also now need to preserve the subdivision loop to
subdivision edge map, which until now was overwritten to store coarse edges
(the map to coarse edges is still preserved).
Also the paint flag stored in the 4th dimension of the loop normal buffer
was not properly set for flat shaded faces, leading to other kind of
artefacts and render issues.
The problem was when the stroke had less weights that the total number of vertex groups.
The API checked the total number of groups, but this is not required because `BKE_defvert_find_index` returns NULL is the vertex group index does not exist.
If creating partial hierarchy overrides (from the outliner e.g.), that
need to create extra overrides, those could be re-created everytime,
leading to very bad situation where there would be several overrides of
the same reference ID in the same override hierarchy.
This fix makes it so that existing overrides in a given hierarchy will
always be re-used in case other overrides are added to the hierarchy.
Replaces old-style memzero-style of call with zero-initializer.
Allows to shorten typical initialization code to a single line:
Object foo = blender:🧬:shallow_zero_initialize<Object>()
This will allow to more easily convert designated initializer
which is often used to fill object with zeroes to the explicit
function calls:
MyDNAStruct foo = {};
will be translated to
MyDNAStruct foo = blender:🧬:shallow_zero_initialize<MyDNAStruct>();
Differential Revision: https://developer.blender.org/D14486
Previously, those methods would destruct and reconstruct
the data structure. While that was more simple in initial
implementation, it has some downsides which are not resolved:
* Already allocated memory is lost. So new memory would have
to be allocated when the data structure is refilled.
* The clearing process itself was slower because it did unnecessary
work.
This simplifies debugging, and can help improve performance
by making it easier for the compiler.
More optimization might still be possible by using `__restrict` in
a few places.
Regression in [0] which removed a special check when tweak events ended.
Add a similar check for drag events that runs drag is disabled in the
main event handling loop.
[0]: 4986f71848
Since e49bf4019b, animation is handled explicitly. Split operator
wasn't updated.
Re-use backup-duplicate-restore animation functions, that other
operators use for splitting.
Regression in [0] error iterating over pose bones which only used the
active-object, also follow the same logic as edit-mode for using the
local-matrix.
[0]: d052169e7e
Regression in [0], also use pad buffer by 1 instead of 2 which is no
longer needed as the trailing slash is no longer added
after allocating the string.
0682af0d63
This is meant to allow using C++ data structures in this file
as a performance improvement. Particularly `Vector` instead
of `ListBase` for `duplilist`. This changes builds on all
platforms on the buildbot.
This is meant to allow using C++ data structures in this file
as a performance improvement. Particularly `Vector` instead
of `ListBase` for `duplilist`. This change builds and passes
tests on all platforms on the buildbot.
This is meant to allow using C++ data structures in this file
as a performance improvement. Particularly `Vector` instead
of `ListBase` for `duplilist`.
Differential Revision: https://developer.blender.org/D14475
If the `Automatic Constraint` modifier was activated while an axis
constraint was already set, the orientation used would be the default
orientation of the mode and not that of the scene.
This was because the `initSelectConstraint` function was not called in
this case and the `Automatic Constraint` mode was enabled by other
indirect means.
So the solution is to call `initSelectConstraint` in either case and
remove these "indirect means" of enabling `Automatic Constraint`.
Contrary to the initial intention (in rB9916e0193c36), `TREDRAW_SOFT`
flag, when isolated, is not cleared in `transformApply` and therefore is
used in the `drawTransformApply` callback which basically recalculates
the `transformation` which finally clears the flag.
So remove the `drawTransformApply` callback so `transformApply` is not
called when unnecessary.
Differential Revision: https://developer.blender.org/D14430
The solution supposedly listed all cases that `absolute grid snapping`
was supported. But it ignored some occasions like: Editing Surface
objects, Texture Space.
List now only the cases where this feature should not be supported.
The Library Overrides display mode is meant to show overridden
properties from the current file only, not library overrides in
data-blocks that just were linked in. The upcoming Hierarchies view mode
for Library Overrides will also display linked in data-blocks that have
overrides in the source file (but not the individual overridden
properties), see T95802.
This was a mistake in the conditional structure introduced in 4b35d6950d
This commit also adds a new type of snap exclusion: `SNAP_NOT_EDITED`.
Thanks to @Ethan1080 for pointing out the error.
In preparation for supporting packing of UDIM tiled textures, this patch
refactors a small portion of image.cc. The refactor should lead to less
duplicate code now and when Tiled images are added in the near future.
This patch is based on the prior work done for D6492 where it was
requested this part be split and can be summarized as follows:
- `load_sequence_single` is removed and merged with `load_image_single`
- `image_load_sequence_file` is removed and merged with `image_load_image_file`
Reviewed By: lukasstockner97
Differential Revision: https://developer.blender.org/D14327
On Windows/MSVC this gives a minor (~20%) speedup presumably due to a faster float/int formatter. On macOS (Xcode13), this gives a massive speedup, since snprintf that is in system libraries ends up spending almost all the time inside some locale-related mutex lock.
The actual exporter code becomes quite a bit smaller too, since it does not have to do any juggling to support std::string arguments, and the buffer handling code is smaller as well.
Windows (VS2022 release build, Ryzen 5950X 32 threads) timings:
- Blender 3.0 splash scene (2.4GB obj): 4.57s -> 3.86s
- Monkey subdivided level 6 (330MB obj): 1.10s -> 0.99s
macOS (Xcode 13 release build, Apple M1Max) timings:
- Blender 3.0 splash scene (2.4GB obj): 21.03s -> 5.52s
- Monkey subdivided level 6 (330MB obj): 3.28s -> 1.20s
Linux (ThreadRipper 3960X 48 threads) timings:
- Blender 3.0 splash scene (2.4GB obj): 10.10s -> 4.40s
- Monkey subdivided level 6 (330MB obj): 2.16s -> 1.37s
The produced obj/mtl files are identical to before.
Reviewed By: Howard Trickey, Dalai Felinto
Differential Revision: https://developer.blender.org/D13998
so a layer can be occluded by the scene instead of always showing in front
---
{F12827163}
Reviewed By: fclem, antoniov
Differential Revision: https://developer.blender.org/D13931
A user asked for this increase. The performance lags when reaching
the upper limit of this number of segments, but if you need that
many segments, I guess you are willing to wait.
The code that eats away faces until you find input faces in
the Constrained Delaunay Triangulation goes too far and crashes
when there are no input faces. In the test case there were input
faces but they only had two vertices, so were all ignored.
In the Blender File display mode of the Outliner, mouse hovering a
"base" element (e.g. "Objects", "Materials", ...) would also highlight
that same base element in other libraries linked into the scene. In fact
operations like (un)collapsing would be applied to both too.
Issue was that we'd always use the listbase containing the data-blocks
from the current main as a way to identify the tree element. So for the
same data-block types we'd use the same listbase pointers. Instead use
the the library pointer + a per library index.
USD requires to be linked with /WHOLEARCHIVE so
the linker won't remove their static initializers.
This strangely has never worked for MSVC since
the flags were set on the LINK_FLAGS property
which is only used to link .dll and .exe files,
given this is a static lib, the flags were not
used, nor did CMake propagate the link directive
to the final targets that did link. Not quite sure
how this has not lead to more problems in the past.
Setting the link directive on the INTERFACE_LINK_OPTIONS
makes cmake do the right thing.
Differential Revision: https://developer.blender.org/D14394
Reviewed by: sybren
For 3.2 USD will be bumped to a newer version with some
slight API changes, however since we cannot simultaneously
land the libs for all platforms as well as these code changes,
we'll have to support both 21.02 and 21.11+ for at least a
short period of time making the code slightly more messy than
it could have been.
Differential Revision: https://developer.blender.org/D14184
Reviewed by: sybren
In order to allow interpolation of integers with a float, add a separate
template parameter for the factor and multiplication types.
Also move some helper constexpr variables to the "base" header
(reversing the dependency to "base" -> "vector").
This also adds a distance function for scalar types, which is
helpful to allow sharing code between vectors and basic types.
Differential Revision: https://developer.blender.org/D14446
This commit implements generic evaluation for Bezier curves (which is
really just linear interpolation, since attributes are not stored on
Bezier handles). For complete parity with the old curve type, we would
have to add options for this (RNA: `Spline.radius_interpolation`),
but it's not clear that we want to do that.
This also adds a generic `interpolate_to_evaluate` utility on curves
that hides the implementation details. Though there is theoretically
a performance cost to that, without some abstraction calling code
would usually be too complex.
Differential Revision: https://developer.blender.org/D14447
This commit adds a utility that returns an array with the number
of curves of every type. One use case for this is detecting whether
to remove handle or NURBS attributes when changing curve types.
It's best to avoid using this when it's not necessary, but sometimes
it can't really be avoided, and having a utility at least makes using
an optimized version simple.
In the future, this information can be cached in the curves runtime.
Differential Revision: https://developer.blender.org/D14448
This patch makes the grease pencil smooth operation symmetric.
It also increases the performance a lot if strong smoothing is
required. Additionally there is an option for the position smooth
operation to keep the shape closer to the original for more iterations.
Since the result differs from the previous algorithm, versioning is used
to change the iterations and factor to match the old result.
Differential Revision: http://developer.blender.org/D14325
Thanks Jacques for finding solution for deprecation warning
which was generated by GCC for constructor.
The rest of the change is related on fixing memaccess warning
which was happening when memset/memcpy was used directly on
the DNA object pointer. Now there are two utility functions
for this:
- blender:🧬:zero_memory
- blender:🧬:copy_memory
This change makes it possible to add implementation of common
C++ methods for DNA structures which helps ensuring unsafe
operations like shallow copy are done explicitly.
For example, creating a shallow copy used to be:
Object temp_object = *input_object;
In the C++ context it was seen like the temp_object is
properly decoupled from the input object, while in the
reality is it not. Now this code becomes:
Object temp_object = blender:🧬:shallow_copy(*input_object);
The copy and move constructor and assignment operators are
now explicitly disabled.
Other than a more explicit resource management this change
also solves a lot of warnings generated by the implicitly
defined copy constructors w.r.t dealing with deprecated fields.
These warnings were generated by Apple Clang when a shallow
object copy was created via implicitly defined copy constructor.
In order to enable C++ methods for DNA structures a newly added
macro `DNA_DEFINE_CXX_METHODS()` is to be used:
tpyedef struct Object {
DNA_DEFINE_CXX_METHODS(Object)
...
} Object;
For the shallow copy use `blender:🧬:shallow_copy()`.
The implementation of the memcpy is hidden via an internal DNA
function to avoid pulling `string.h` into every DNA header.
This means that the solution does not affect on the headers
dependencies.
---
Ideally `DNA_shallow_copy` would be defined in a more explicit
header, but don;t think we have a suitable one already. Maybe
we can introduce `DNA_access.h` ?
Differential Revision: https://developer.blender.org/D14427
This reverts commit 8c44793228.
Apparently, this generated a lot of warnings in GCC.
Didn't find a quick solution and is it not something I want to be
trading between (more quiet Clang in an expense of less quiet GCC).
Will re-iterate on the patch are re-commit it.
This change makes it possible to add implementation of common
C++ methods for DNA structures which helps ensuring unsafe
operations like shallow copy are done explicitly.
For example, creating a shallow copy used to be:
Object temp_object = *input_object;
In the C++ context it was seen like the temp_object is
properly decoupled from the input object, while in the
reality is it not. Now this code becomes:
Object temp_object = blender:🧬:shallow_copy(*input_object);
The copy and move constructor and assignment operators are
now explicitly disabled.
Other than a more explicit resource management this change
also solves a lot of warnings generated by the implicitly
defined copy constructors w.r.t dealing with deprecated fields.
These warnings were generated by Apple Clang when a shallow
object copy was created via implicitly defined copy constructor.
In order to enable C++ methods for DNA structures a newly added
macro `DNA_DEFINE_CXX_METHODS()` is to be used:
tpyedef struct Object {
DNA_DEFINE_CXX_METHODS(Object)
...
} Object;
For the shallow copy use `blender:🧬:shallow_copy()`.
The implementation of the memcpy is hidden via an internal DNA
function to avoid pulling `string.h` into every DNA header.
This means that the solution does not affect on the headers
dependencies.
---
Ideally `DNA_shallow_copy` would be defined in a more explicit
header, but don;t think we have a suitable one already. Maybe
we can introduce `DNA_access.h` ?
Differential Revision: https://developer.blender.org/D14427
Clear the nurbs basis cache dirty flag when its evaluation finishes.
Remove an incorrect assert that the evaluated size couldn't be zero.
It can, when `check_valid_size_and_order` returns false.
Add a function to retrieve the points for an index range of curves,
and move "ensuring" the offsets to a separate function, since it's
often nicer to call that if you don't need the result span immediately.
When boolean fields are evaluated and used as selections, we create
a vector of indices. This process is currently single-threaded, but
226f0c4fef added a more optimized multi-threaded version
of this process. It's simple to use this in the field evaluator.
I tested this with the set position node and a random
value node set to boolean mode on a Ryzen 2700x:
| | Before | After | Improvement |
| 10% Selected | 40.5 ms | 29.0 ms | 1.4x |
| 90% Selected | 115 ms | 45.3 ms | 2.5x |
In the future there could be a specialized version for non-span
virtual array selections that uses `materialize` to lower virtual
call overhead.
Differential Revision: https://developer.blender.org/D14436
Replace comparisons of FT_Error against 0 with FT_Err_Ok instead.
See D14052 for more details.
Differential Revision: https://developer.blender.org/D14052
Reviewed by Campbell Barton
Currently only supports single image frames (no animation possible).
If quality slider is set to 100 then lossless compression will be used,
otherwise lossy compression is used.
Gives about 35% reduction of filesize save when re-saving splash screens with lossless
compression.
Also saves much faster, up to 15x faster than PNG with a better compression ratio as a plus.
Note, this is currently left disabled until we have WebP libs (see T95206)
For testing precompiled libs can be downloaded from Google:
https://storage.googleapis.com/downloads.webmproject.org/releases/webp/index.html
Differential Revision: https://developer.blender.org/D1598
Each time the user clicks the viewport 2 sets of engine views are
created. Each set is currently composed of 8 view objects, each of size
592 bytes.
Because space is not reserved in the vector that holds them, several
unnecessary re-allocation/copy cycles occur as the vector resizes and
the total allocation load is 8880 bytes. This happens twice.
Reduce to just the allocations necessary and with exactly 4736 bytes
allocated for each set
- Before: 8 allocations and 8 deallocations totaling 17760 bytes
- After: 2 allocations and 2 deallocations totaling 9472 bytes
Reviewed By: fclem, jbakker
Differential Revision: https://developer.blender.org/D13782
In the library overrides mode, in some situations there would be empty
base elements like "Collections" or "Objects". Don't show them, it's
confusing wihout use. Code just failed to consider that case.
All the buttons in the Library Overrides display mode would be shown in cyan,
indicating that they are library overrides. Given that this is solely what this
display mode is about, the indicator is just redundant, confusing (why are the
buttons purple?) and looks weird.
Part of T95802.
Reviewed by: Bastien Montagne
Differential Revision: https://developer.blender.org/D14416
Fix provided by Piotr Makal in patch D14204.
This patch fixes volume grid duplication which was occurring during
importing USD files. This was caused by calling BKE_volume_grid_add
twice per grid (excluding 'density' grid) for the same Volume
object: (1) in USDVolumeReader::read_object_data and (2) later in
BKE_volume_load.
Differential Revision: https://developer.blender.org/D14204
When creating a new window from a duplicated area - by shift-dragging
on corner action zones - on the Windows platform the resulting window
is initially unresponsive. This patch fixes this by releasing the parent
window's mouse capture.
See D14085 for more details.
Differential Revision: https://developer.blender.org/D14085
Reviewed by Ray Molenkamp
There is a dedicated Library Override display mode now, and showing
these elsewhere just adds noise and makes the code problematic to
maintain (since the same element hierarchy will be used in two entirely
different contexts). The corresponding filter settings are removed too.
Part of T95802.
Reviewed by: Bastien Montagne
Differential Revision: https://developer.blender.org/D14411
Instead of having the "Current File" and then the individual libraries
containing indirect library overrides in the Library Overrides display
mode, only show what's in the current file. Agreement was that this
isn't very useful in this view, we may want to add it to the Hierarchy
view though (see T95802).
Part of T95802.
Also expands the top level ID type items ("Objects", "Materials", ...)
by default. See D14410 for details.
Reviewed by: Bastien Montagne
Differential Revision: https://developer.blender.org/D14410
Was a mistake in coordinate handling: need to take possible offset
from the sliding area.
Mouse us still not fully perfectly follows changes of the tilt slide
area, but that is a separate issue.
This new modifier creates a shape known as envelope. It connects all
points that are n points apart. There is also a mode which fits a
single stroke to the envelope shape that is determined by that rule.
For more details, refer to the patch.
Reviewed By: NicksBest, antoniov, frogstomp, mendio
Differential Revision: http://developer.blender.org/D14341
This implements the same interpolation method as for bevel weights
now for vertex and edge creases as well to improve the flexibility.
Differential Revision: http://developer.blender.org/D14170
Essentially, we only allow deletion of hierarchy roots of liboverrides,
when hierarchy deletion option is enabled.
Also add some checking code in the generic, non-object/collection ID
delete code, to prevent any deletion of liboverrides that would be part
of a hierarchy.
Object mode selection does a kind of cycling that excludes the active
selected object. This is separate from regular selection cycling which
is enabled when clicking multiple times without moving the cursor.
This has the down-side that clicking on an object to drag it always
selects the object behind it (in the case of overlapping objects).
Since object mode selection is fundamental functionality, this is
exposed as an experimental preference for user feedback & testing.
See T96752 for details.
In my opinion Scale Thickness feels similar to Only Locations and was missing from the pie menu in grease pencil edit mode, so here I added it.
Differential Revision: https://developer.blender.org/D12530
This patch maximizes the possible bounds for the dash and
gap parameters in the dot dash modifier. This makes e.g.
chopping up lines without gaps possible.
Differential Revision: http://developer.blender.org/D14428
This was already as a preference for the tweak tool,
this preference enables the option for all selection
in the 3D view & UV editor.
This extends on changes from T96544.
Always prioritizing bones caused pose-objects to be selected in object
mode even if they were behind other objects.
Now prioritizing pose bones is limited to pose mode.
When cycling through objects select the nearest first
instead of using the order of object-bases in the view_layer.
This matches how pose selection works.
This is useful to save time manually averaging many timing results.
The minimum is included because often it can be more stable than an
average, and it can help to expose calls from other contexts with lower
times that would make the average useless.
Differential Revision: https://developer.blender.org/D14417
In a test file from T96282, this commit reduces the runtime of the
delete geometry node from 82 ms to 23 ms, a 3.6x improvement.
Writing to vertex groups in other cases should be faster too.
The largest improvement comes from not writing a new weight
of zero if the vertex is not in the group. This mirrors the behavior
of custom data interpolation in `layerInterp_mdeformvert`.
Other improvements come from using `set_all` for writing
output attributes and implementing that method for vertex groups.
I also implemented `materialize` methods. Though I didn't obverse
an improvement from this, I think it's best to remove virtual method
call overhead where it's simple to do so.
The test file for the delete geometry node needs to be updated.
These methods could be parallelized too, but better to do that later.
Differential Revision: https://developer.blender.org/D14420
Rename "size" variables and functions to use "num" instead,
based on T85728 (though this doesn't apply to simple C++
containers, it applies here). Rename "range" to "points" in
some functions, so be more specific.
Differential Revision: https://developer.blender.org/D14431
There are some filenames where the UDIM pattern guessing would fail
unnecessarily. The user can fix these up afterwards but it would be
nicer if they would detect properly in the first place.
Examples:
`test.1001.ver0023.png` would guess wrong since it uses the image
sequence detection code which finds the first sequence from the end. It
would guess `filename.1001.ver<UDIM>.png`
`uv-test.u1_v2.png` would fail detection due to a bug in the processing.
Make this much more reliable and add tests for the most important tile
related get/set/detection functions.
Differential Revision: https://developer.blender.org/D14320
Basic testing on windows only so far. Will need some testing on Linux as well
when the Linux enablement patch is ready.
Does not enable Vega APUs yet (which would be gfx902 or gfx90c).
Differential Revision: https://developer.blender.org/D14432
For render image buffers to be acquired, a lock must be provided. Also
fixed wrong usage of release, it must always be called regardless if the
returned image buffer is NULL.
Explicit template specialization has to happen outside of class
definition (some compilers are more lenient). Since it is not possible to
specialize the method without also specializing the enclosing class for
all of its possible types, the method is moved outside of the class, and
specialized there.
This adds a const qualifier to some code path in the Alembic and USD
importers. More could be added elsewhere. This change is done as it will
be required when GeometrySets are supported and helps keeping diff noise
in the patch to a bare minimum.
ViewMapIO.h is only included in Controller.cpp init_options().
However, g_models_path and g_flags set here are never used elsewhere.
Therefore, ViewMapIO files can be deleted without affecting anything.
Differential Revision: https://developer.blender.org/D14423
Use templates to optimize the CPU texture sampler to interpolate using
float for single component datatypes instead of using float4 for all types.
Differential Revision: https://developer.blender.org/D14424
New supported features:
* 3D/spherical brush that samples a good position on the curves.
* Falloff.
The custom falloff curve mapping is not yet available in the ui because that
requires some more ui reorganization. This is better done when we have
a better understanding of what settings we need exactly.
Currently, the depth of the 3d brush is only sampled once per stroke, when
first pressing LMB. Sometimes it is expected that the depth of the brush can
change within a single brush. However, implementing that in a good way
is not straight forward and might need additional options. Therefore that
will be handled separately. Some experimentation results are in D14376.
Ref T96445.
Differential Revision: https://developer.blender.org/D14376
rBc1909770e7f192574ea62449dd14b4254637e604 introduced "PXR_LIB_PREFIX" for building the
dependencies, so only makes sense to use the same name in the Hydra render delegate CMake too
This patch adds a Hydra render delegate to Cycles, allowing Cycles to be used for rendering
in applications that provide a Hydra viewport. The implementation was written from scratch
against Cycles X, for integration into the Blender repository to make it possible to continue
developing it in step with the rest of Cycles. For this purpose it follows the style of the rest of
the Cycles code and can be built with a CMake option
(`WITH_CYCLES_HYDRA_RENDER_DELEGATE=1`) similar to the existing standalone version
of Cycles.
Since Hydra render delegates need to be built against the exact USD version and other
dependencies as the target application is using, this is intended to be built separate from
Blender (`WITH_BLENDER=0` CMake option) and with support for library versions different
from what Blender is using. As such the CMake build scripts for Windows had to be modified
slightly, so that the Cycles Hydra render delegate can e.g. be built with MSVC 2017 again
even though Blender requires MSVC 2019 now, and it's possible to specify custom paths to
the USD SDK etc. The codebase supports building against the latest USD release 22.03 and all
the way back to USD 20.08 (with some limitations).
Reviewed By: brecht, LazyDodo
Differential Revision: https://developer.blender.org/D14398
If the `Automatic Constraint` modifier was activated while an axis
constraint was already set, the orientation used would be the default
orientation of the mode and not that of the scene.
This was because the `initSelectConstraint` function was not called in
this case and the `Automatic Constraint` mode was enabled by other
indirect means
So the solution is to call `initSelectConstraint` in either case and
remove these "indirect means" of enabling `Automatic Constraint`.
This moves `MOD_meshsequencecache.c` to C++ and fixes compile warnings
introduced from the change. This uses C++ style casts, as well as
`nullptr` instead of `NULL`.
This will allow to output `GeometrySets` from the modifier, which is C++.
Differential Revision: https://developer.blender.org/D13662
This enables the configuration to specify aliases for compatibility with other
configurations.
When a colorspace name is saved in a.blend, that is the alias of a colorspace
in the current configuration, it will show the main colorspace from the
configuration in the user interface and Python API instead.
Loading & saving the .blend file does not make any changes to the stored name,
so as to not make hidden modifications. Only when setting the property again
will the alias name be overwritten by the main colorspace name.
Fixes T96049
Differential Revision: https://developer.blender.org/D14419
The Output Properties > Output panel now has a Color Management subpanel to
override scene settings. When set to Override instead of Follow Scene, there
are settings to:
* For OpenEXR, choose a (linear) colorspace for RGBA passes
* For other file formats, use different display/view/look/exposure/gamma
These settings affect animation render output, image save of renders and the
compositor file output node. Additionally, the image save operator and
compositor file output nodes also support overriding color management.
Includes some layout changes to the relevant panels to accomdate the new
settings and to improve consistency. Ideally subpanels would be used to better
organize these settings, however nodes and operators don't currently support
creating subpanels.
Differential Revision: https://developer.blender.org/D14402
When using GradingPrimaryTransform the generated GLSL code fails to compile. The actual issue is
inside OCIO (https://github.com/AcademySoftwareFoundation/OpenColorIO/issues/1603).
The reason is that unset clamping values are rendered out as `inf`, which isn't recognizable
by GLSL.
The issue is worked around by defining a default for `inf`.
Reviewed By: brecht
Maniphest Tasks: T96502
Differential Revision: https://developer.blender.org/D14425
When converting ghost keys to Blender's event system:
- All keys that aren't part of the GHOST_TKey enum map to EVENT_NONE
(ignored), note that it's an internal error if the value of key isn't
a known value.
- Modify the switch statement so any missing members of GHOST_TKey
warn at compile time (GCC & Clang only).
- GHOST_kKeyUnknown maps to EVT_UNKNOWNKEY.
We could ignore this key, changing can be evaluated separately.
Avoid adding events with their type set to EVENT_NONE as these
can't be categorized usefully (keyboard/mouse/NDOF ... etc),
and add an extra case that needs to be accounted for.
Adding these events seems to be an unintentional change from [0],
these keys used to be ignored in 2.4x.
[0]: a1c8543f2a
This reverts commit e55f4657f7.
It's not intended to support assigning shortcuts to this operator,
which could only work for built-in keying sets caused warnings to be
reported warnings when exporting key-maps.
Prefer D14289, preventing users running into this problem to begin with.
- Follow the same conventions as the 3D viewport for UV selection
(using SelectPick_Params internally).
- Use WM_operator_properties_mouse_select for selection properties.
Vertices were not drawn properly as the logic for mapped mesh was used
in the BMesh case.
Edge display would ignore subdivided edges which would come from coarse
edges when setting display flags.
The checks for calling outliner flushing didn't account for
entering pose mode for the first time or that pose-bone selection
can also change the object selection.
Resolve by recording what changed and refresh accordingly.
Also de-duplicate calls to DEG_id_tag_update.
- Document parameters.
- Add code-comments.
- Remove some historic/unhelpful code-comments.
- Rename argument names that were ambiguous
(object was a boolean for e.g.).
- Move `gpu` picking into an allocated struct which is only
allocated & used when using GPU picking.
- Move variable declarations after menu picking has been handled.
If the mouse is not hovering the window, there is no active region. This is a
valid state, but the UI-list filter operator didn't account for that case.
Tweaks:
- Increase horizontal padding for the buttons from 1 point to 2, looked like an
unintentional placement error before.
Fixes:
- Missing horizontal padding for array buttons
- Small gap between separator line and right column when using a high interface
scale
- Properly center buttons vertically.
The preview does not work well with deferred render result pixels
allocation: it breaks the refresh and requires to toggle current
panels.
Since there is no tiled rendering for previews we don't save any
memory by deferring pixels allocations, so do it for the render
result during the render result creation.
Differential Revision: https://developer.blender.org/D14414
Connecting to some sockets of a few nodes via the drag link search
would fail and trigger an assert, because the picked socket wasn't
available. This was due to some sockets only being available with
certain settings.
This patch fixes these cases by adding the availability conditions of
the socket to the node declaration with the `make_available` method
or manually adding a `node_link_gather_search` function.
Differential Revision: https://developer.blender.org/D14283
Split retrieval of translated text for the "invalid" messages for NURBS
curves from the actual calculation, which is a lower-level function.
Also fixes an issue where "At least two points required" would always
display in the "Active Spline" panel.
Differential Revision: https://developer.blender.org/D14315
A mistake in the mesh normal refactor caused the wrong mesh to
be used when calculating normals with a shape key's deformation.
This commit fixes the normal calculation by using the correct mesh,
with just adjusted vertex positions, and calculating the results
directly into the result arrays when possible. This completely avoids
the need to make a local copy of the mesh, which makes sense,
since the only thing that changes is the vertex positions.
Differential Revision: https://developer.blender.org/D14317
These boolean options are passed as uint, but they would be
easier to understand if using bool.
Differential Revision: https://developer.blender.org/D14405
Currently there is a "calc_face_normal" argument to mesh to bmesh
conversion, but vertex normals had always implicitly inherited whatever
dirty state the mesh input's vertex normals were in. Probably they were
most often assumed to not be dirty, but this was never really correct in
the general case.
Ever since the refactor to move vertex normals out of mesh vertices,
cfa53e0fbe, the copying logic has been explicit: copy the
normals when they are not dirty. But it turns out that more control is
needed, and sometimes normals should be calculated for the resulting
BMesh.
This commit adds an option to the conversion to calculate vertex
normals, true by default. In almost all places except the decimate
and edge split modifiers, I just copied the value of the
"calc_face_normals" argument.
Differential Revision: https://developer.blender.org/D14406
Make a copy of ImageFormatData that contains the effective color management
settings, and pass that along to the various functions. This will make it
possible to add more complex logic later.
For compositing nodes, passing along view and display settings through
many functions made it harder to add additional settings, so just get those
from the scene now.
Differential Revision: https://developer.blender.org/D14401
Metal shading language follows the C++ 14 standard and in some cases requires a greater level of explicitness than GLSL. There are also some small language differences:
- Explicit type-casts (C++ requirements)
- Explicit constant values (C++ requirements, e.g. floating point values using 0.0 instead of 0).
- Metal/OpenGL compatibility paths
- GLSL Function prototypes
- Explicit accessors for vector types when sampling textures.
Authored by Apple: Michael Parkin-White
Ref T96261
Reviewed By: fclem
Maniphest Tasks: T96261
Differential Revision: https://developer.blender.org/D14378
Adding WITH_METAL option to CMAKE to guard compilation for macOS only. Implemented stub METALBackend to mirror GPUBackend interface and added capabilities initialisation, along with API initialisation paths.
Global rendering coordination commands added to backend with GPU_render_begin and GPU_render_end() commands globally wrapping GPU work. This is required for Metal to ensure temporary resources are generated within an NSAutoReleasePool and freed accordingly.
Authored by Apple: Michael Parkin-White, Vil Harvey, Marco Giordano, Michael Jones, Morteza Mostajabodaveh, Jason Fielder
Ref T96261
Reviewed By: fclem
Maniphest Tasks: T96261
Differential Revision: https://developer.blender.org/D14293
Caused by rBc0bd240ad0a1.
To avoid crash, make boolean value false if active object data is NULL.
Should be backported to 2.93 LTS and 3.1 corrective releases.
- No need to store is_pose_mode, check the object_mode flag instead.
- Remove redundant pose-mode check which now skips object selection.
- Remove disabled grease pencil cursor toggling,
since I couldn't manage to reproduce a situation where the cursor
failed to update - also, as there are other places the active object
can change this would need a more general solution anyway.
Selecting that opens a menu is now returns early. Handling the menu
selection in-line made this function more difficult to follow.
Also split out selecting an object by it's center into it's own function.
- Rename baseCount to bone_count, was copy-pasted from object code.
- Also rename baseCount to base_count for object selection,
following snake case naming conventions.
- Use int instead of short for counters, as there is no reason to use
short ints.
A 7 year old commit, 2ec00ea0c1, used incorrect indexing for
the optional array of precomputed poly normals. Apparently that code
path was never used, or this issue would have been discovered earlier.
Recent changes calculate normals on a temporary mesh and use those
for the "low-res" layer, meaning the precomputed path was always taken.
Since 3b6ee8cee7, a list of vertex groups cannot be retrieved
from curve objects for merging because curve objects do not support
vertex groups. Previously the empty list on the object was returned.
Only mesh objects are supported for the caps.
Old python exporter in 3.0 and earlier ordered faces by material,
but the new C++ exporter in 3.1+ did not, and was just writing them
in whatever is the order of the mesh data structure.
This mostly does not cause problems, except in some apps e.g.
Procreate -- for large enough meshes, this lack of
"order by material" (which ends up having more usemtl lines)
ends up creating more mesh subsets than necessary inside Procreate.
The change is not computationally heavy, e.g. exporting 6-level
subdivided Monkey mesh goes 1085ms -> 1105ms on my machine.
Reviewed By: @howardt
Differential Revision: https://developer.blender.org/D14368
This adds a new Add brush for the new curves object type in sculpt mode.
The brush is used to insert new curves (typically hair) on the surface object.
Supported features:
* Add single curve exactly at the cursor position when `Add Amount` is 1.
* Front faces only.
* Independent interpolate shape and interpolate length settings.
* Smooth and flat shading affects curve shape interpolation.
* Spherical and projection brush.
This also adds the `surface_triangle_index` and `surface_triangle_coordinate`
attributes. Those store information about what position on the surface each
added curve is attached to:
* `surface_triangle_index` (`int`): Index of the internal triangle that a curve
is attached to. `-1` when the curve is not attached to the surface.
* `surface_triangle_coordinate` (`float2`): First two numbers of a barycentric
coordinate that reference a specific position within the triangle.
Ref T96444.
Differential Revision: https://developer.blender.org/D14340
Also fixes missing code to read/write/free/copy color management settings
in various places. This can't be set through the UI currently, but still
should be handled consistently.
CPU code for cubic interpolation with clip texture extension only performed
texture interpolation inside the range of [0,1]. As a result, even though the
volume's color is sampled using cubic interpolation, the boundary is not
being interpolated. The GPU appears was interpolating samples that span the
clip boundary softening the edge, which the CPU now does also.
This commit also includes refactoring of 2D and 3D texture sampling in
preparation of adding new extension modes.
Differential Revision: https://developer.blender.org/D14295
Allow the user to set an anisotropic filtering setting below the implementation-defined value of `GL_MAX_TEXTURE_MAX_ANISOTROPY_EXT`.
This bug-fix is also needed for 2.93 LTS.
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D14392
This patch fixes T96655, bloom crashing Eevee.
The error occurs because rB472fc0c55848b2e2d428cfb4f7debb80a4e12081 added `vec3 safe_color(vec3 c)` to `common_math_lib.glsl`.
However, `vec3 safe_color(vec3 c)` already exists in `effect_bloom_frag.glsl`.
This means `vec3 safe_color(vec3 c)` is duplicated within `common_math_lib.glsl` and `effect_bloom_frag.glsl`.
{F12938060 size=full}
The duplicate code in `effect_bloom_frag.glsl` can be removed since it's no longer needed.
(I checked the remaining methods, there shouldn't be any additional duplicate code)
Reviewed By: fclem
Maniphest Tasks: T96655
Differential Revision: https://developer.blender.org/D14396
Consider switching to rendered shading type as a request to start
rendering, without requiring to un-pause.
This minimizes amount of clicks needed to start rendering after
viewport was paused once, and then shading mode got changed.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D14244
Hello,
I saw that this revision was stalled for a few months so I tried to update it.
https://developer.blender.org/D10995
I added a function that adds a preview on the first connected input of the file output node.
I removed the preview on the single layer format
Thanks
Reviewed By: #compositing, jbakker
Maniphest Tasks: T84815
Differential Revision: https://developer.blender.org/D14219
This is a general issue exposed by moving from tweak to click-drag
events [0], however this bug would have existed for both click &
click-drag events beforehand.
Since [1] the following behavior could occur:
- Click-drag the cursor away from the button.
- Leaving the button would flag it as disabled.
- The disabled button would then break
causing the event to be considered handled.
- Once handled no click / click-drag action would be tested.
The bug would only happen if the cursor left the button before the drag
threshold was reached which tended to happen with an UI-scale 2 or more.
Or with an increased drag threshold.
Revert [1] (fix for T78503), which is no longer needed since as of [2].
[0]: 4986f71848
[1]: 6f96dd8576
[2]: 87c13ac68c
- Use set-style printing of modifier flags instead of booleans.
- Include event.flag.
- Print on a single line (so output can be more easily filtered).
In some cases value of cursor motion events was set from the last event
written to the wmWindow.eventstate. Avoid potential errors by ensuring
cursor motion events always have their value set to KM_NOTHING.
This is patch D14349 from Aras Pranckevicius.
The logic in the code was _completely different_ from the documentation
and what the python exporter in 3.0 did. The new code assumed that
"export material groups" meant "append material name to the object name",
and was only ever kicking in when the "export object groups" option was
also checked. But the proper behavior (as in 3.0 exporter & the online docs),
is to emit g objectname_materialname before each usemtl line. Which is something entirely else.
This is patch D14347 from Aras Pranckevicius.
Instead of scaling "the scene" (i.e. transform vertices by object matrix,
then multiply by scale factor), it was instead first applying the scale
factor in local space, and then transforming by the object matrix.
This is quite a huge cleanup. Making use of the `common_gpencil_lib.glsl`
to share more codes and use more consistent codestyle.
The gpencil engine specifics are now out of the `gpencil_vertex()`
function making it easier to add more features.
There should be no regression as all workarounds are kept as is.
Some C headers might define the typedefs of the enum themselves.
Even if they are guarded by preprocessor `#if`, our enum preprocessor
has no idea of the validity of the statement. So we just bypass
if there is a typedef just before any `enum` keywords.
Note that the typedef matching is quite fragile.
This library contains the needed functions to render GPencil object
geometry. Centralizing these will make it possible for other draw
engines (EEVEE, Overlay) to reuse the same vertex shader code and
possibly the same fragment rejection methods.
This adds a simple and more manageable temp texture behavior.
The texture is garanteed to be available only between the acquire/release pair.
This makes the same engine able to reuse the textures and even overlap the acquire
& release calls.
- Add name support to storage buffers
- Delete view functions for TextureFromPool
- Add support for different size acquire and assert on mulitple acquire
- Allow multiple release
This applies patch D14343 from Aras Pranckevicius, with a description:
The new 3.1+ OBJ exporter did not have correct logic when faced with
non-uniform & mirrored (negative on odd number of axes) object scale:
- Normals were not transformed correctly (should use inverse transpose of the matrix),
and were not normalized,
- Face order was not "flipped" when transform has negative scale on odd number of axes
(visible when using "face orientation" viewport overlay).
When filtering File Browser items by name, use entry's "name" field as
well as the "relpath" field since they can vary.
See D13940 for details.
Differential Revision: https://developer.blender.org/D13940
Reviewed by Bastien Montagne
Windows IME: Fix duplicated initial character when entering numbers
while in Chinese full width character mode.
See D14354 for more details.
Differential Revision: https://developer.blender.org/D14354
Reviewed by Brecht Van Lommel
Previously, the conversion was done manually for a fixed set of types.
Now, there is a more general utility that can be used in other contexts
(outside of geometry nodes attribute processing) as well.
The fast_float library provides fast header-only implementations for the C++ from_chars
functions for `float` and `double` types. These functions convert ASCII strings representing
decimal values (e.g., `1.3e10`) into binary types. We provide exact rounding (including
round to even). In our experience, these `fast_float` functions many times faster than comparable number-parsing functions from existing C++ standard libraries.
Specifically, `fast_float` provides the following two functions with a C++17-like syntax (the library itself only requires C++11):
```C++
from_chars_result from_chars(const char* first, const char* last, float& value, ...);
from_chars_result from_chars(const char* first, const char* last, double& value, ...);
```
The return type (`from_chars_result`) is defined as the struct:
```C++
struct from_chars_result {
const char* ptr;
std::errc ec;
};
```
It parses the character sequence [first,last) for a number. It parses floating-point numbers expecting
a locale-independent format equivalent to the C++17 from_chars function.
The resulting floating-point value is the closest floating-point values (using either float or double),
using the "round to even" convention for values that would otherwise fall right in-between two values.
That is, we provide exact parsing according to the IEEE standard.
Given a successful parse, the pointer (`ptr`) in the returned value is set to point right after the
parsed number, and the `value` referenced is set to the parsed value. In case of error, the returned
`ec` contains a representative error, otherwise the default (`std::errc()`) value is stored.
The implementation does not throw and does not allocate memory (e.g., with `new` or `malloc`).
It will parse infinity and nan values.
Example:
``` C++
#include "fast_float/fast_float.h"
#include <iostream>
int main() {
const std::string input = "3.1416 xyz ";
double result;
auto answer = fast_float::from_chars(input.data(), input.data()+input.size(), result);
std::cout << "parsed the number " << result << std::endl;
return EXIT_SUCCESS;
}
```
Like the C++17 standard, the `fast_float::from_chars` functions take an optional last argument of
the type `fast_float::chars_format`. It is a bitset value: we check whether
`fmt & fast_float::chars_format::fixed` and `fmt & fast_float::chars_format::scientific` are set
to determine whether we allow the fixed point and scientific notation respectively.
The default is `fast_float::chars_format::general` which allows both `fixed` and `scientific`.
The library seeks to follow the C++17 (see [20.19.3](http://eel.is/c++draft/charconv.from.chars).(7.1)) specification.
* The `from_chars` function does not skip leading white-space characters.
* [A leading `+` sign](https://en.cppreference.com/w/cpp/utility/from_chars) is forbidden.
* It is generally impossible to represent a decimal value exactly as binary floating-point number (`float` and `double` types). We seek the nearest value. We round to an even mantissa when we are in-between two binary floating-point numbers.
Furthermore, we have the following restrictions:
* We only support `float` and `double` types at this time.
* We only support the decimal format: we do not support hexadecimal strings.
* For values that are either very large or very small (e.g., `1e9999`), we represent it using the infinity or negative infinity value.
We support Visual Studio, macOS, Linux, freeBSD. We support big and little endian. We support 32-bit and 64-bit systems.
## Using commas as decimal separator
The C++ standard stipulate that `from_chars` has to be locale-independent. In
particular, the decimal separator has to be the period (`.`). However,
some users still want to use the `fast_float` library with in a locale-dependent
manner. Using a separate function called `from_chars_advanced`, we allow the users
to pass a `parse_options` instance which contains a custom decimal separator (e.g.,
std::cout << "parsed the number " << result << std::endl;
return EXIT_SUCCESS;
}
```
## Reference
- Daniel Lemire, [Number Parsing at a Gigabyte per Second](https://arxiv.org/abs/2101.11408), Software: Pratice and Experience 51 (8), 2021.
## Other programming languages
- [There is an R binding](https://github.com/eddelbuettel/rcppfastfloat) called `rcppfastfloat`.
- [There is a Rust port of the fast_float library](https://github.com/aldanor/fast-float-rust/) called `fast-float-rust`.
- [There is a Java port of the fast_float library](https://github.com/wrandelshofer/FastDoubleParser) called `FastDoubleParser`.
- [There is a C# port of the fast_float library](https://github.com/CarlVerret/csFastFloat) called `csFastFloat`.
## Relation With Other Work
The fastfloat algorithm is part of the [LLVM standard libraries](https://github.com/llvm/llvm-project/commit/87c016078ad72c46505461e4ff8bfa04819fe7ba).
The fast_float library provides a performance similar to that of the [fast_double_parser](https://github.com/lemire/fast_double_parser) library but using an updated algorithm reworked from the ground up, and while offering an API more in line with the expectations of C++ programmers. The fast_double_parser library is part of the [Microsoft LightGBM machine-learning framework](https://github.com/microsoft/LightGBM).
## Users
The fast_float library is used by [Apache Arrow](https://github.com/apache/arrow/pull/8494) where it multiplied the number parsing speed by two or three times. It is also used by [Yandex ClickHouse](https://github.com/ClickHouse/ClickHouse) and by [Google Jsonnet](https://github.com/google/jsonnet).
## How fast is it?
It can parse random floating-point numbers at a speed of 1 GB/s on some systems. We find that it is often twice as fast as the best available competitor, and many times faster than many standard-library implementations.
('MULTIPLE_IMPORTANCE_SAMPLING',"Multiple Importance Sampling","Multiple importance sampling is used to combine direct light contributions from next-event estimation and forward path tracing",0),
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
