OctaneRender for Maya 2018.1.1-14.26 [OBSOLETE] Tweet

[JimStar]

Otoy® is proud to announce the availability of a new OctaneRender™ for Maya® Version 2018.1
The world's fastest and most feature-rich unbiased render engine that integrates completely into Autodesk® Maya®.

Maya® Version Requirements

This release will work with Maya® Versions 2016.5, 2017, 2018 and 2019 64-bit on Windows, Linux and OSX operating systems.

To use this version, you need a GPU of compute model 3.0 or higher. Support for Fermi GPUs (like GTX 4xx and 5xx) has been dropped. Also make sure you use a driver of version 419.17 or higher and if you are working on Mac OS that have a CUDA driver installed that supports at least CUDA 9.1.


Changelog and the installers for Octane Standalone are here:
OctaneRender Standalone 2018.1.1


Changes since version 14.25:

• Implemented the support of full Ornatrix hair-set rendering in Octane viewport-rendering session. You need to enable the new checkbox in Octane's viewport settings dialog. Still some visibility artifacts possible in some cases, but it's another Ornatrix issue Ephere is already aware of...
• Fixed a bug that lead to transparency inversion in a new fluid shading-translation feature.
• Added "invert" attribute to Octane fluid's translation settings.
• Some efficiency improvements in the new fluid shading-translation.

=================================
Subscription Studio Downloads
=================================

OctaneRender for Maya Studio [Windows]

OctaneRender for Maya Studio [Linux]

OctaneRender for Maya Studio [MacOS]


=================================
Subscription Enterprise Downloads
=================================

OctaneRender for Maya Enterprise [Windows]

OctaneRender for Maya Enterprise [Linux]

OctaneRender for Maya Enterprise [MacOS]


=================================
DEMO Downloads
=================================

OctaneRender for Maya DEMO [Windows]

OctaneRender for Maya DEMO [Linux]

OctaneRender for Maya DEMO [MacOS]

[JimStar]

There is nothing special in the simple ovdbWrite export tool's code compared to what Octane translation does. When comparing with the ovdbWrite please use the -t flag with it (along with "-e -v" flags). Otherwise it does NOT translate the textured shading (just skips it and takes density channel instead directly, which is quick in Maya API), in which case it does NOT use the Maya's slow fluid-sampling, and thus does NOT do the same thing that Octane is forced to do. Octane translation may be slightly slower than simple export tools for other reasons, but not dramatically.

[calus]

There is nothing special in the simple ovdbWrite export tool's code compared to what Octane translation does. When comparing with the ovdbWrite please use the -t flag with it (along with "-e -v" flags). Otherwise it does NOT translate the textured shading (just skips it and takes density channel instead directly, which is quick in Maya API), in which case it does NOT use the Maya's slow fluid-sampling, and thus does NOT do the same thing that Octane is forced to do. Octane translation may be slightly slower than simple export tools for other reasons, but not dramatically.

Obviously I compared using ovdbWrite texture flag and even checked the exported VDB was the expected result, but still octane plugin was at a lot slower.

indeed this is good news if Octane use the same method meaning there might be a bug somewhere else but I will run more advanced comparison tests to find the culprit.

[calus]

- In this release Fluid translation time is much better, but not very stable.
For a textured fluid with 100 voxel resolution, at first render in a new Maya session it takes 15s to translate to IPR,
but then if I change a fluide attribute, translation takes 30 s, or sometimes 1 mn.
For information Arnold for the same grid have no noticeable translation time, so I guess it is using it's own procedurals avoiding the slow Maya API part.

- Maybe this is not relative but I noticed the number of voxels in info is always wrong always twice the number it should be, so for a 100x100x100 voxel fluid grid the result in octane info is 2 million voxels instead of 1 million as expected.

- "opacity" is reversed by default so I need to enable the new"voxel invert" switch to get correct result but it's not stable at all during IPR, sometime doesn't work at all.

I need lot more testing...