This results in a speedup if the capture attribute is only needed
under specific circumstances (e.g. when a switch further down the
line is true). Previously, the input field was always evaluated.
Differential Revision: https://developer.blender.org/D15018
* Port over new code tables from Cycles
* Convert Rec.709 to scene linear for lookup table.
* Move code for wavelength and blackbody to IMB so they can access the
required transforms, which are not in blenlib.
* Remove clamping from blackbody shader to bypass the texture read.
Since it's variable now easiest to just always read from the texture
than pass additional parameters.
* Fold XYZ to RGB conversion into the wavelength table.
Ref T68926
* Rename ambiguous rgb to scene_linear in some places
* Precompute matrices to directly go to scene instead of through XYZ
* Make function signatures more consistent
This patch changes the Text Box Width socket to always have that label
instead of switching to "Max Width" when Overflow mode is picked.
Bug report: T97060
Differential Revision: https://developer.blender.org/D14909
Subtracting one from the evaluated index could make the index -1.
That was only necessary for Bezier curves due to the specifics of
the "bezier_evaluated_offsets".
Geometry Nodes (new) icon. So far we were using the generic node-tree
icon for geometry nodes, not anymore.
The new icon is composed of 4 spheres that is a reference to the
original pebbles demo. Scattering points was also the turning point for
the project (which originally was focusing on dynamic effects), and to
this day is one of the first steps for everything procedural such as
hair.
Note that the modifier icon is still showing as white in the outliner.
The alternative is to be blue everywhere.
Patch review and feedback by Hans Goudey.
Icon creation in collaboration with Pablo Vazquez.
This patches rewrites the GPU shaders of curve nodes for easier future
development. This is a non-functional change. The new code avoids code
duplication by moving common code into BKE curve mapping functions. It
also avoids ambiguous data embedding into the gradient vectors that are
passed to vectors and reduces the size of uniforms uploaded to the
shader by avoiding redundancies.
This is needed in preparation for the viewport compositor, which will
utilize and extend this implementation.
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D14689
There are fancier possibilities for improvements, like taking ownership
of existing arrays in some cases, but this patch takes a simpler brute
force approach for now.
The first change is to move from the previous loop to using the new
`materialize_compressed_to_uninitialized` method on virtual arrays,
which adds only the selected values to the output. That is a nice
improvement in some cases, corresponding to the "Without Threading"
column in the chart below.
The next change is to call that function in parallel on slices of
the output. To avoid generating too much code, we can avoid
templating based on the type and devirtualizing completely.
The test input is a 4 million point grid, generated by the grid
primitive node. Color and 2D vector attributes were also transferred
to the points.
| Test | Before | Final No Threading | Final | Change |
| --------- | ------ | ------------------ | ------ | ------ |
| All Verts | 209 ms | 186 ms | 170 ms | 0.8x |
| 1% | 148 ms | 143 ms | 133 ms | 0.9x |
| All Faces | 326 ms | 303 ms | 87 ms | 0.27x |
| 1% Faces | 70 ms | 68 ms | 34 ms | 0.49x |
Differential Revision: https://developer.blender.org/D14661
This adds an input to the Subdivision node to specify a field to use
for controling vertex creases. Common code with edge creasing was
extracted into utility functions to avoid redundancy.
Differential Revision: https://developer.blender.org/D14199
Inspired by D12936 and D12929, this patch adds general purpose
"Combine Color" and "Separate Color" nodes to Geometry, Compositor,
Shader and Texture nodes.
- Within Geometry Nodes, it replaces the existing "Combine RGB" and
"Separate RGB" nodes.
- Within Compositor Nodes, it replaces the existing
"Combine RGBA/HSVA/YCbCrA/YUVA" and "Separate RGBA/HSVA/YCbCrA/YUVA"
nodes.
- Within Texture Nodes, it replaces the existing "Combine RGBA" and
"Separate RGBA" nodes.
- Within Shader Nodes, it replaces the existing "Combine RGB/HSV" and
"Separate RGB/HSV" nodes.
Python addons have not been updated to the new nodes yet.
**New shader code**
In node_color.h, color.h and gpu_shader_material_color_util.glsl,
missing methods hsl_to_rgb and rgb_to_hsl are added by directly
converting existing C code. They always produce the same result.
**Old code**
As requested by T96219, old nodes still exist but are not displayed in
the add menu. This means Python scripts can still create them as usual.
Otherwise, versioning replaces the old nodes with the new nodes when
opening .blend files.
Differential Revision: https://developer.blender.org/D14034
The issue was that the `NodeTreeRef` acceleration data structure was
rebuild much more often than necessary. That happened because the
Map Range node accidentally tagged the node tree for change even
though it did not actually change.
Differential Revision: https://developer.blender.org/D14842
Use `src` and `dst` instead of less common variable names,
less redundant logic, simpler use of const, and "typename"
for template arguments instead of "class".
Overwriting UV map or vertex color data in Geometry nodes will move the
layers to another CustomData channel, and as such, will make attribute
lookup fail from the UVMap and Vertex Color nodes in EEVEE as the
CustomDataType will also be modified (i.e. no longer `CD_MTFACE` or
`CD_MCOL`).
As discussed in T93179, the solution is to use `CD_PROP_AUTO_FROM_NAME`
so that the render engine is able to find the attributes. This also makes
EEVEE emulate Cycles behaviour in this regard. `attr_load_uv` and
`attr_load_color` are also removed in favor of the generic attribute
API in the various GLSL shaders.
Although `CD_PROP_AUTO_FROM_NAME` is now used even for UV maps, the
active UV map is still used in case the attribute name is empty, to
preserve the old behavior.
Differential Revision: https://developer.blender.org/D13730
This commit implements copying of materials and material indices from
all of the boolean node's input meshes. The materials are added to the
final mesh in the order that they appear when looking through the
materials of the input meshes in the same order of the multi-socket
input node.
All material remapping is done with mesh-level materials. Object-level
materials are not considered, since the meshes don't come from objects.
Merging all materials rather than just the materials on the first mesh
requires a change to the boolean-mesh conversion. This subtly changes
the behavior for object linked materials, but in a good way I think;
now the material remap arrays are respected no matter the number
of materials on the first mesh input.
Differential Revision: https://developer.blender.org/D14788
At least on the Mac, `std::sort` sometimes passes the same value in the
`a` and `b` parameters.
The `true` return is only for cases where `a` is less than `b`.
Goals:
* Better high level control over where devirtualization occurs. There is always
a trade-off between performance and compile-time/binary-size.
* Simplify using array devirtualization.
* Better performance for cases where devirtualization wasn't used before.
Many geometry nodes accept fields as inputs. Internally, that means that the
execution functions have to accept so called "virtual arrays" as inputs. Those
can be e.g. actual arrays, just single values, or lazily computed arrays.
Due to these different possible virtual arrays implementations, access to
individual elements is slower than it would be if everything was just a normal
array (access does through a virtual function call). For more complex execution
functions, this overhead does not matter, but for small functions (like a simple
addition) it very much does. The virtual function call also prevents the compiler
from doing some optimizations (e.g. loop unrolling and inserting simd instructions).
The solution is to "devirtualize" the virtual arrays for small functions where the
overhead is measurable. Essentially, the function is generated many times with
different array types as input. Then there is a run-time dispatch that calls the
best implementation. We have been doing devirtualization in e.g. math nodes
for a long time already. This patch just generalizes the concept and makes it
easier to control. It also makes it easier to investigate the different trade-offs
when it comes to devirtualization.
Nodes that we've optimized using devirtualization before didn't get a speedup.
However, a couple of nodes are using devirtualization now, that didn't before.
Those got a 2-4x speedup in common cases.
* Map Range
* Random Value
* Switch
* Combine XYZ
Differential Revision: https://developer.blender.org/D14628
