- Avoid using geometry sets from a different abstraction level
- Deduplicate basic attribute copying propagation code
- Allow more use of implicit sharing when data arrays are unchanged
- Optimize for when a point cloud delete selection is empty
- Handle face corners generically for "only face" case
Pass the curves and points to keep instead of delete. In the same test
file as the previous commit, this gave an increase from 50 to 60 FPS
when deleting curves.
Goals of this refactor:
* Reduce memory consumption of `IndexMask`. The old `IndexMask` uses an
`int64_t` for each index which is more than necessary in pretty much all
practical cases currently. Using `int32_t` might still become limiting
in the future in case we use this to index e.g. byte buffers larger than
a few gigabytes. We also don't want to template `IndexMask`, because
that would cause a split in the "ecosystem", or everything would have to
be implemented twice or templated.
* Allow for more multi-threading. The old `IndexMask` contains a single
array. This is generally good but has the problem that it is hard to fill
from multiple-threads when the final size is not known from the beginning.
This is commonly the case when e.g. converting an array of bool to an
index mask. Currently, this kind of code only runs on a single thread.
* Allow for efficient set operations like join, intersect and difference.
It should be possible to multi-thread those operations.
* It should be possible to iterate over an `IndexMask` very efficiently.
The most important part of that is to avoid all memory access when iterating
over continuous ranges. For some core nodes (e.g. math nodes), we generate
optimized code for the cases of irregular index masks and simple index ranges.
To achieve these goals, a few compromises had to made:
* Slicing of the mask (at specific indices) and random element access is
`O(log #indices)` now, but with a low constant factor. It should be possible
to split a mask into n approximately equally sized parts in `O(n)` though,
making the time per split `O(1)`.
* Using range-based for loops does not work well when iterating over a nested
data structure like the new `IndexMask`. Therefor, `foreach_*` functions with
callbacks have to be used. To avoid extra code complexity at the call site,
the `foreach_*` methods support multi-threading out of the box.
The new data structure splits an `IndexMask` into an arbitrary number of ordered
`IndexMaskSegment`. Each segment can contain at most `2^14 = 16384` indices. The
indices within a segment are stored as `int16_t`. Each segment has an additional
`int64_t` offset which allows storing arbitrary `int64_t` indices. This approach
has the main benefits that segments can be processed/constructed individually on
multiple threads without a serial bottleneck. Also it reduces the memory
requirements significantly.
For more details see comments in `BLI_index_mask.hh`.
I did a few tests to verify that the data structure generally improves
performance and does not cause regressions:
* Our field evaluation benchmarks take about as much as before. This is to be
expected because we already made sure that e.g. add node evaluation is
vectorized. The important thing here is to check that changes to the way we
iterate over the indices still allows for auto-vectorization.
* Memory usage by a mask is about 1/4 of what it was before in the average case.
That's mainly caused by the switch from `int64_t` to `int16_t` for indices.
In the worst case, the memory requirements can be larger when there are many
indices that are very far away. However, when they are far away from each other,
that indicates that there aren't many indices in total. In common cases, memory
usage can be way lower than 1/4 of before, because sub-ranges use static memory.
* For some more specific numbers I benchmarked `IndexMask::from_bools` in
`index_mask_from_selection` on 10.000.000 elements at various probabilities for
`true` at every index:
```
Probability Old New
0 4.6 ms 0.8 ms
0.001 5.1 ms 1.3 ms
0.2 8.4 ms 1.8 ms
0.5 15.3 ms 3.0 ms
0.8 20.1 ms 3.0 ms
0.999 25.1 ms 1.7 ms
1 13.5 ms 1.1 ms
```
Pull Request: https://projects.blender.org/blender/blender/pulls/104629
- Samplerate -> Sample rate: should be two words.
- "Falloff type the feather": typo.
- JPEG, OpenJPEG and JPEG 2000 are the official spellings of the
respective projects.
- "... boundary of image(including ...": missing space.
- "Points in . direction (cannot be changed ...)":
Plural, it is a collection of multiple points. Also do not use
contraction for "cannot".
- The Bevel modifier's "Only Vertices" option was replaced by a
Vertices mode in 2.90.
- "Metaball Types": affects one object, should be singular.
- "Metaball data-block to defined blobby surfaces": typo.
- "Wire Size": this option has nothing to do with wireframes, I suppose
it's an old terminology.
- "... (for negative speed.)": remove trailing period.
- "Smooth factor effect": the prop describes a factor for an effect,
not an effect for a factor.
- "... assigned to their vertices(ensures ...": missing space.
- "... used when faces have the ObColor mode enabled": ObColor is not
used anywhere else in the UI (since Blender 2.50).
- "Effect Children": typo -> Affect.
- "Distort Min/Max": copy-pasted from another pair of properties.
- "... dismiss menu on release.(in 1/100ths of sec)": replace period
with space.
- "resolution": field names should be capitalized.
Pull Request: https://projects.blender.org/blender/blender/pulls/108227
Combine the newer less efficient C++ implementations and the older
less convenient C functions. The maps now contain one large array of
indices, split into groups by a separate array of offset indices.
Though performance of creating the maps is relatively unchanged, the
new implementation uses 4 bytes less per source element than the C
maps, and 20 bytes less than the newer C++ functions (which also
had more overhead with larger N-gons). The usage syntax is simpler
than the C functions as well.
The reduced memory usage is helpful for when these maps are cached
in the near future. It will also allow sharing the offsets between
maps for different domains like vertex to corner and vertex to face.
A simple `GroupedSpan` class is introduced to make accessing the
topology maps much simpler. It combines offset indices and a separate
span, splitting it into chunks in an efficient way.
Pull Request: https://projects.blender.org/blender/blender/pulls/107861
In dd32dac60f, the "A" and "B" input socket from the Mix node were
disambiguated, so as not to confuse them with Alpha and Blue.
These messages are used in other nodes and elsewhere in the same
sense, so this commit adds translation contexts to these occurrences
as well.
Pull Request: https://projects.blender.org/blender/blender/pulls/108051
The Vector Math node's label has a special function used to compute
its label based on the currently selected operation. This operation,
like that of the Math node, is extracted using the "NodeTree"
translation context.
Therefore, in this function it must also use the same context to
translate the label.
In addition, a few node types can have an "Unknown" label if there is
a problem with their internal state (operation, filter type, blending
type). This message can also be translated.
Pull Request: https://projects.blender.org/blender/blender/pulls/108029
Store bevel weights in two new named float attributes:
- `bevel_weight_vert`
- `bevel_weight_edge`
These attributes are naming conventions. Blender doesn't enforce
their data type or domain at all, but some editing features and
modifiers use the hard-coded name. Eventually those tools should
become more generic, but this is a simple change to allow more
flexibility in the meantime.
The largest user-visible changes are that the attributes populate the
attribute list, and are propagated by geometry nodes. The method of
removing this data is now the attribute list as well.
This is a breaking change. Forward compatibility is not preserved, and
the vertex and edge `bevel_weight` properties are removed. Python API
users are expected to use the attribute API to get and set the values.
Fixes#106949
Pull Request: https://projects.blender.org/blender/blender/pulls/108023
If there are no loose vertices or edges, it's basically free to
propagate that information to the result and save calculating
it later in case it's necessary. I observed a peformance increase
from 3.6 to 4.1 FPS when extruding a 1 million face grid.
Used to be https://archive.blender.org/developer/D17123.
Internally these are already using the same code path anyways, there's no point in maintaining two distinct nodes.
The obvious approach would be to add Anisotropy controls to the Glossy BSDF node and remove the Anisotropic BSDF node. However, that would break forward compability, since older Blender versions don't know how to handle the Anisotropy input on the Glossy BSDF node.
Therefore, this commit technically removes the Glossy BSDF node, uses versioning to replace them with an Anisotropic BSDF node, and renames that node to "Glossy BSDF".
That way, when you open a new file in an older version, all the nodes show up as Anisotropic BSDF nodes and render correctly.
This is a bit ugly internally since we need to preserve the old `idname` which now no longer matches the UI name, but that's not too bad.
Also removes the "Sharp" distribution option and replaces it with GGX, sets Roughness to zero and disconnects any input to the Roughness socket.
Pull Request: https://projects.blender.org/blender/blender/pulls/104445
Since 44e4f077a9 and related commits, geometry nodes doesn't
try to hide the difference between real geometry data and instances from
the user. Other nodes were updated to only support real geometry, but
the "Mesh Boolean" node was never updated and still implicitly gathered
all the instances. This commit removes the special instance behavior in the
boolean node and adds realize instances nodes to keep existing behavior
in most cases. Typically this doesn't make a difference in the result,
though it could in the union mode for instance inputs. Shifting more of
the work to realizing instances should generally be better for
performance, since it's much faster.
The goal is to make it more explicit and centralized operation to
assign and steal buffer data, with proper ownership tracking.
The buffers and ownership flags are wrapped into their dedicated
structures now.
There should be no functional changes currently, it is a preparation
for allowing implicit sharing of the ImBuf buffers. Additionally, in
the future it is possible to more buffer-specific information (such
as color space) next to the buffer data itself. It is also possible
to clean up the allocation flags (IB_rect, ...) to give them more
clear naming and not have stored in the ImBuf->flags as they are only
needed for allocation.
The most dangerous part of this change is the change of byte buffer
data from `int*` to `uint8_t*`. In a lot of cases the byte buffer was
cast to `uchar*`, so those casts are now gone. But some code is
operating on `int*` so now there are casts in there. In practice this
should be fine, since we only support 64bit platforms, so allocations
are aligned. The real things to watch out for here is the fact that
allocation and offsetting from the byte buffer now need an explicit 4
channel multiplier.
Once everything is C++ it will be possible to simplify public
functions even further.
Pull Request: https://projects.blender.org/blender/blender/pulls/107609
This commit adds a new experimental node "Sample Volume".
It's purpose is to be able to sample values from a volume grid given
a position field.
The 'Grid' input accepts a named attribute input with the name
of the volume grid to sample. Values will be implicitly converted
from the grid type to the specified value type.
Pull Request: #107656
The previous two commits introduced new regexes to extract node socket
names and descriptions automatically from the i18n Python module,
including socket labels.
This commit removes the extraction macros from all node files since
they are now redundant.
Pull Request: https://projects.blender.org/blender/blender/pulls/107258
In some languages, the A and B sockets need to be disambiguated
because because they can have different letters depending on context.
In Spanish, A is used for B (Blue -> Azul), while the greek letter
alpha is used for A, to disambiguate Azul / alpha.
Pull Request: https://projects.blender.org/blender/blender/pulls/107706
See: https://projects.blender.org/blender/blender/issues/103343
Changes:
1. Added `BKE_node.hh` file. New file includes old one.
2. Functions moved to new file. Redundant `(void)`, `struct` are removed.
3. All cpp includes replaced from `.h` on `.hh`.
4. Everything in `BKE_node.hh` is on `blender::bke` namespace.
5. All implementation functions moved in namespace.
6. Function names (`BKE_node_*`) changed to `blender::bke::node_*`.
7. `eNodeSizePreset` now is a class, with renamed items.
Pull Request: https://projects.blender.org/blender/blender/pulls/107790
This patch implements the Convert Color Space node for the realtime
compositor. A custom OCIO GpuShaderCreator was implemented to use the
ShaderCreateInfo in constructing the OCIO GPU processor shader. That
shader is then cached inside the cache manager and is invalidated when
the OCIO configuration changes.
Pull Request: https://projects.blender.org/blender/blender/pulls/107878
Generally render engines can do subframe mixing themselves, but the
purpose of subframe mixing in the simulation output node is to support
higher quality motion blur with bakes when there are topology-changing
operations after the simulation output node. Linear mixing can fill the
gaps while maintaining lower memory usage.
All point/instance domain attributes are mixed, but mixing is only
supported when the domain size is unchanged or when an `id` attribute
gives a mapping between elements. Theoretically it may be possible, but
nested instance geometry is not mixed in this commit due to the
difficulty of finding matching geometries across arbitrary instance
hierarchy changes. Attributes that are completely unchanged are ignored
using implicit sharing for better performance.
Pull Request: https://projects.blender.org/blender/blender/pulls/107599
This makes the Blender binary 350 KB smaller. The largest change comes
from using `FunctionRef` instead of a template when gathering indices to
mix in the extrude node (which has no performance cost). The rest of the
change comes from consolidating uses of code generation for all
attribute types. This brings us a bit further in the direction of
unifying attribute propagation.
Pull Request: https://projects.blender.org/blender/blender/pulls/107823
Adds an input node Signed Distance that can
be used to refer to the distance-grid of an SDF volume.
Right now it's experimental and can be enabled by enabling
"New Volume Nodes" in the preferences.
It returns a float AttributeFieldInput called 'distance' which
can tell Volume nodes which grid to get.
See #103248.
This patch implements the Plane Track Deform node for the realtime
compositor. The implementation is mostly similar to the Corner Pin node,
but it is generalized to multiple homography matrices to take motion
blur into account.
Pull Request: https://projects.blender.org/blender/blender/pulls/107811
In vertex extrusion mode, when there are no edge attributes we can avoid
creating a topology map meant for mixing old values. This makes
simulation caching for a curl noise demo file from user Higgsas 10%
faster (from 44s to 40s to calculate 230 frames).
