OctaneRender™ 4.0 for LightWave™ - Beta build 4.00.0.14 RC4

[juanjgon]

Hi! This is the third beta release of the OctaneRender™ 4.0 for LightWave™ plugin version 4.00.0.14, available for LightWave 11.6, 2015.3 and 2018.0 Windows x64 and OSX. This build is compiled using the Octane 4.00 RC4 core, including some of its new features, like the lights linking or the new Universal material. The new UDIM node will be available in the next plugin build.

As with all experimental and test releases, please do not use this build for production work, and please note that we still classify this as an experimental release.

Important notes:
- NVIDIA Driver version 387 or higher is required on Windows and Linux, and any CUDA driver for Mac should work that supports CUDA 9.1.
- Octane no longer supports Fermi GPUs.
- When starting Octane for the first time, you'll see a window to download the cuDNN lib. Occasionally when we upgrade cuDNN, you may see this again at a later date. You can simply click Download and Octane will download the file needed. If you have any trouble with this process please let us know. If you need to, you can manually download the file and copy it into place next to the plugin .p file

For introductory material on V4, please see the viewtopic.php?f=33&t=66013 and viewtopic.php?f=33&t=68283


Release 4.00.0.14 regular license
https://render.otoy.com/customerdownloa ... _Win64.zip
https://render.otoy.com/customerdownloa ... ve_OSX.zip

Release 4.00.0.14 subscription license
https://render.otoy.com/customerdownloa ... n64_vr.zip
https://render.otoy.com/customerdownloa ... OSX_vr.zip

Standalone version for the network rendering nodes
https://render.otoy.com/customerdownloa ... C4_win.exe
https://render.otoy.com/customerdownloa ... _macos.dmg


=============================================
RELEASE 4.00.0.14
=============================================

* PLUGIN FEATURES:
- New option in the render passes node to save the noisy beauty image file or EXR layer when the denoiser is enabled.

* OCTANE FEATURES:
- New Universal material node.
- Light linking and global light exclusion. New object properties plugin params and render target options to configure the light masks.
- Added ability to capture direct and/or indirect light for light passes.
- Added near clip plane support for Z depth pass.
- Added options in texture and blackbody emission nodes to keep emissive power constant when applying uniform scaling (if surface brightness is disabled).
- New volumetric AI denoiser.

* OCTANE BUGS FIXED:
- Fixed incorrect behaviour when more than 64 render passes are enabled, which also caused a lot of error messages.
- Fixed texture preview.
- Fixed a crash when loading very small images.
- Fixed light linking passes 7 and 8.
- Fixed volume rendering when there are multiple volumes in the scene.
- Fixed some specular material differences from 3.08.
- Fixed denoise results when affect alpha option is enabled .
- Fixed AI Light with animated triangles in the scene.
- Fixed random colour texture.
- Fixed a performance issue when using direct light kernel with a medium or volume where absorption is very low.
- Improve OSL compiler performance.
- Fixed denoiser output when using affect alpha without an alpha channel.
- Fixed a case where network render GPU count was incorrect.

- Fixed artefacts when using high roughness on brdfs other than Octane.
- Fixed displacement crash on some meshes.
- Fixed Beckmann sampling for the specular case.
- Fixed some artefacts in compressed textures.
- Fixed some memory not being deallocated sometimes when using AI Light.
- Some performance optimizations for AI Light processing
- Speed up loading of .hdr files
- Improved how Octane deals with running out of memory, which fixes some cases where caching fails and then a render failure occurs.
- A small volume rendering performance improvement of between 5-10%.
- Fixed a significant performance regression from 3.08, where scenes with hundreds of unique meshes rendered significantly slower.
- Render speed improvements - in general v4 is now faster than v3, and info channel kernels are between 15-25% faster.
- Fixed crash caused by invalid primitives with emission.
- We no longer distribute the cuDNN library with Octane - when octane first loads, we download cuDNN if it has changed.

- Fixed denoiser blend option not working properly when using hot pixel filter.
- Fixed render failure when denoiser output render pass is enabled, and the denoiser is disabled in the imager node.
- Fixed layer shadow pass.
- Increased the maximum number of OSL nodes in one texture node tree from 500 to approximately 30,000.
- Fixed crash when rendering some meshes that have multiple materials where one or more have displacement.
- Fixed a bug where alpha values were written into the output image buffer when the alpha channel was disabled.
- Fixed two problems with the render layer shadow pass: shadows when there was no emitter, and shadow strengths being incorrect.
- Fixed crashes when rendering simple scenes that contain an emitter.
- Fixed render failure on some simple meshes.
- Fixed render failure on some meshes that happens after caching is complete.
- Fixed incorrect memory consumption reporting sometimes.
- Fixed a possible hang when using AI Light and modifying other settings.
- Fixed a crash when using AI Light in a scene with no emitters.
- Fixed render failures when AI Light processing is interrupted with either emission changes or other changes in the scene.
- Fixed some extra memory consumption when switching from a scene with emitters to one that has none.
- Fixed a case where device memory stats returns total memory as unavailable memory.
- Fixed a crash when using AI Light if all lights in the scene have a sample rate of zero.
- Fixed a case where SSS data is accidentally added into the transmission pass.
- Fixed AI Light crash if an emitter is moved immediately before mesh caching begins.
- Fixed clay mode for universal materials.
- Fixed normal passes to record incorrectly sometimes.
- Fixed the export of raw buffers when the resolution is not a multiple of 4.
- Fixed a crash and incorrect motion blur that happens sometimes when making changes in placement nodes.
- Fixed handling of int[2] and int[3] input parameters for OSL shaders.
- Fixed a very random render failure.
- Fixed AI Light update for Network Render.
- Fixed baking camera.
- Fixed a GGX sampling error.
- Nishita sky: Fixed some very bright sunlight from Nishita sky model when the sun is just below the horizon.
- Planetary environment: Improved appearance of star field
- Planetary environment: Fixed rendering of sunset at high altitudes
- Added an approximation for ozone absorption to the planetary environment.
- Fixed a crash on the roughness info pass.
- Fixed a crash when attempting to render hairs with zero vertices.
- Fixed rendering of meshes with an opacity map inside the environment medium.
- Fixed crash due to (unsupported) AVX instructions in the denoiser.
- Fixed crash due to failing to page-lock the result buffer of the denoiser.
- Fixed crash in the denoiser if the film resolution is not a multiple of 4.
- Fixed render failure if the scene doesn't contain instances.
- Fixed a network render crash when AI light is disabled.
- Fixed incorrect UV mapping of on-the-fly compressed image textures.
- Fixed artefacts in some image compresison formats (when done on-the-fly).
- Fixed missing Volta support on OSX.


Thanks,
-Juanjo

[BorderLine]

Thanks JuanJo !

[Lewis]

Great, Thanks for implementing separate Denoiser Pass Juan, that will make my day .

[LightwaveGuru]

Holy! YEAH!

snip lwguru

[vipvip]

Thanks a lot.... impressive fix/add list !!

[FrankPooleFloating]

Thanks so very much Juanjo! Great job, as always! You have been a busy boy, for sure.

Now that you have this in the bag, is there any way you can revisit the idea (which I believe you said was totally doable) of making the Gradient2 node dynamic -- where we can add/subtract to it -- and not have 36 keys by default, with no ability to remove unneeded keys?... Please-o-please? That dang thing takes up soooooooo much valuable real estate.

I think that when we discussed this before, I might have floated the idea that if dynamic length of node was not possible, maybe you could give us an additional version with only 8-10 keys... I don't think I have ever used more than 10 keys myself...

I never use the regular Gradient Texture node, but maybe that could get this as well?...

Edit: Ye gods!!! 28 total inputs with sixteen different texture inputs in Universal Material!!! Holy crap! I knew UM was going to have lots of new bells and whistles.. but daaaaaang...

[stef]

Thanks Juanjo !

It's fantastic what you do for Lightwave !
Great continuation to you, just a little things,
Image+Color correct+Environment = no heading !

Best regards.
_

[pixym]

Just curious : is there a plan for having Vectron in host app ?

[promity]

It seems there is a problem with the matte function -

The function does not work correctly, in addition there are reflections on the material, which in principle can not contain them.

P.S. I understood the reason - if at least one light source has a disabled shadow option - the matte surface is not displayed correctly.

[promity]

Hmm. Constantly falls on scenes with animated instances (trees). Less often falls when using a denoiser.