This is the third alpha release of v3

We split the v3 alpha release notes into two posts: This one, which provides the download links and information specific to this particular release and a features post describing and discussing the new features and changes of 3.00. The features post will be continuously updated until we reach the stable version 3.00.
Alpha and beta version disclaimer
In contrast to v2 we want to do the alpha and beta testing publicly, to get more feedback earlier and to have changes tested more thoroughly. The plan is to let all users with a valid v2 license use the v3 alpha and beta versions until we decide that the beta period has ended and we release the first stable v3. After that you will have to upgrade to v3 or will not be able to use any v3 versions (including alpha and beta versions) anymore. We don't know yet, when that will be, but we will give you a warning a few weeks in advance so you can prepare.
Since there is a higher likelihood of severe bugs in alpha and beta versions you shouldn't use them for production work. It would be nice if you could test some production scenes with the alpha and beta versions, but you should definitely not rely on the stability of those versions to finish your projects.
In the Standalone you should be able to load v3 scenes in v2, losing only features provided in v3 and maybe some converted settings (like the new "mid level" in displacement nodes). With the plugins this isn't necessarily the case, which means that you should not create any plugin projects with v3 alpha/beta if you are not 100% sure that you want to upgrade to v3 or that you don't mind losing these projects when the beta period ends.
Manual
We don't have an updated manual for v3 yet. That's partially due to the fact that we are currently overhauling the documentation web site, which needs to be finished until we can publish an updated manual for v3. We will let you know when the v3 manual is available until then, please refer to the v3 new features thread and its discussion.
CUDA requirements
We switched to CUDA 7, which means on Windows and Linux you may need to update your NVIDIA graphics driver to a current version which supports at least CUDA 7.0. On Mac OSX you may have to update the CUDA driver for Mac to a current version supporting at least CUDA 7.0.
Changes since 3.00 alpha 2
- Upgraded OpenSubDiv to version 3, which should be more robust than the previous version. We had to completely rewrite the glue code (between Octane and the library) which means that there is a higher likelyhood of bugs and issues with subdivision surfaces.
- Added support for large amounts of geometry, i.e. you can render now scenes with more than 19.6 million triangles. Please be aware though that it still needs to fit into GPU memory.
- Added saving and loading of render pass export configurations (see File -> Preferences... -> Export).
- As a precursor of the planned Octane plugin API, we started to overhaul the Standalone UI, to make it more flexible. As a first result, most dialogs are now resizable and more consistent. This is still very early days and under heavy construction...
- Added support to upload animated image textures / volumes to ORC.
- Volume emission works now different depending on which emission node is used: The temperature channel of a volume will modulate the temperature if a black body emission node is used and the emission power if a texture emission node is used.
- The absorption and scatter channels default to a grey colour now so that when loading volumes, they don't come up blank immediately. To make it work with the specular materials, the medium pin of a specular material will stay empty initially.
- Exposure and hot pixel removal also work with the baking camera now.
- Improved baking edge noise removal.
- Improved compilation of acceleration structure for texture baking.
- Fixed out-of-core memory setup on Linux.
- Fixed a crash when you render a material or texture with an empty input linker connected with the projection pin.
- Fixed a crash when switching from a multiple instances scene to a single instance scene with baking camera.
- Fixed a bug which would render the inverse normal when a position is specified for the baking camera.
- Fixed rendering of volumes with small gaps.
- Fixed writing of ORBX packages with wrong permissions on Linux.
- It's now easier to save all render passes at once in Lua:
octane.render.saveRenderPassesMultiExr()
andoctane.render.saveRenderPasses()
acceptnil
as the render passes list to save everything with the default names.
Known issues
Region rendering is still broken and we are working to bring it back as soon as possible. Unfortunately it requires a total rewrite due to the film buffer changes.
Also render priority is currently not implemented due to the kernel overhaul. Actually, since we are now launching much smaller kernels more often, we don't see the need for this functionality anymore and want to remove it, but aren't sure if there are still cases where we would need a throttling of the kernel calls. Please let us know, in case you run into issues like a blocked/laggy UI.
Baking camera still doesn't work with meshes with displacement textures.
OctaneRender Standalone downloads
Windows
- 64-bit commercial version (installer)
- 64-bit commercial version (ZIP archive)
Mac OS X
- 64-bit commercial version (DMG image)
Linux
- 64-bit commercial version (ZIP archive)
Photoshop workflow plugin download
You can download and install the updated plugin either via the Adobe Exchange platform or you can download it from the link below and manually install it:
Windows and Mac OS X
- Photoshop extension (ZXP archive)
Happy rendering,
Your OTOY NZ Team