arnold-usd
This repository contains a set of components and tools to use the Arnold renderer with Pixar's Universal Scene Description. Notably, the following components are included:
Hydra render delegate
Arnold procedural for USD
Schemas to describe an Arnold scene in USD
Contributions are welcome! Please make sure to read the contribution guidelines.
Please follow the building instructions. To use the components, provided you installed in <arnold-usd_dir>
, set the following environment variables:
Add <arnold-usd_dir>/lib/python
to PYTHONPATH
for the Python schema bindings.
Add <arnold-usd_dir>/plugin
to PXR_PLUGINPATH_NAME
for the Hydra render delegate and the Node Registry plugin.
Add <arnold-usd_dir>/lib/usd
to PXR_PLUGINPATH_NAME
for the USD schemas.
Add <arnold-usd_dir>/lib
to LD_LIBRARY_PATH
on Linux, PATH
on Windows and DYLD_LIBRARY_PATH
on Mac.
The render delegate currently supports the following features:
RPrim Support
All primvars are supported, st/uv is accessible through the built-in uv attribute on the mesh
Support for the displayColor primvar
Subdivision settings
Mesh
Volume
Points
SPrim Support
Distant Light
Sphere Light
Disk Light
Rect Light
Cylinder Light
Dome Light
Arnold shaders are supported, the info:id
attribute is used to determine the shader type
UsdPreviewSurface is translated to Arnold shaders
Materials
Lights
BPrim Support
Render Buffer
OpenVDB Asset
Point Instancer, including nesting of Point Instancers
Selection in USD View and other applications using the primId
AOV
Displaying the Color, Depth and PrimID AOVs
Motion Blur
Deformation
Transformation
Render Settings via the Render Delegate
Sampling parameters
Threading parameters
Ignore parameters
Profiling and logging parameters
Switching between CPU and GPU mode seamlessly
Default values are configurable through environment variables for most of these parameters
Limitations
Crash on linux at startup with usdview : Currently, the arnold library needs to be ld-preloaded to avoid the crash, e.g. LD_PRELOAD=/path_to_arnold/bin/libai.so usdview scene.usda
No motion blur support for the Point Instancer attributes
Can’t preview arbitrary primvar AOVs
No basisCurves
No field3d volume grids
Not all the parameters are accessible through the render settings
Texture Cache size
Texture generation parameters (automip, autotile)
No normal maps on the UsdPreviewSurface
Only converging renders are supported (ie. it’s not possible to block the viewport until the render finishes)
No HdExtComputation and UsdSkel computation via the render delegate
No of physical camera parameters
No coordsys support
The Node Registry plugin supports the current features:
Registering Sdr nodes for every built-in shader and custom shader
Setting up the asset URI either to <built-in>
or to the path of the shader library providing the shader.
Creating all parameters.
Limitations
No nodes registered for
Shapes
Lights
Filters
Drivers
No node is registered for the options node
Metadata is not converted for any node
The procedural supports the following features:
USD shapes
UsdGeomMesh
UsdGeomCurves
UsdGeomPoints
UsdGeomCube
UsdGeomSphere
UsdGeomCone
UsdGeomCylinder
primvars are translated as user data
USD Lights
UsdLuxDistantLight
UsdLuxDomeLight
UsdLuxDiskLight
UsdLuxSphereLight
UsdLuxRectLight
UsdLuxGeometryLight
Support for textured lights (dome and rectangle)
USD native shaders
UsdPreviewSurface
UsdPrimVar*
UsdUVTexture
Arnold shaders supported as UsdShade nodes (where info:id gives the shader type)
Support for any additional Arnold parameter in USD nodes (e.g. attribute arnold:subdiv_iterations
in a UsdGeomMesh)
Support for any Arnold node type (e.g. USD type ArnoldSetParameter gets rendered as arnold set_parameter
node)
Support for multi-threaded parsing of a USD file
LimitationsCurrently unsupported:
Nurbs
Point Instancer
Cameras
Luma Pictures' usd-arnold
RodeoFX's OpenWalter
Ben Asher
Sebastien Blaineau-Ortega
Chad Dombrova
Guillaume Laforge
Julian Hodgson
Thiago Ize
Pal Mezei
Paul Molodowitch
Nathan Rusch
Frederic Servant
上一篇:cogitareTask
下一篇:ignite-ui
还没有评论,说两句吧!
热门资源
Keras-ResNeXt
Keras ResNeXt Implementation of ResNeXt models...
seetafaceJNI
项目介绍 基于中科院seetaface2进行封装的JAVA...
spark-corenlp
This package wraps Stanford CoreNLP annotators ...
capsnet-with-caps...
CapsNet with capsule-wise convolution Project ...
inferno-boilerplate
This is a very basic boilerplate example for pe...
智能在线
400-630-6780
聆听.建议反馈
E-mail: support@tusaishared.com