We are excited to bring you the first experimental release of Octane 2020.2.
As with all experimental builds, we advise to not to use this release for production purposes. We can't guarantee that scenes saved with this version will be compatible in future releases, and we anticipate there being several maintenance releases.
We look forward to your feedback on this release. You can test 2020.2 XB1 if you have a valid Octane subscription.
Make sure to use a NVIDIA Studio driver with version at least 451 on Windows or 450 on Linux to use this version. No driver updates are required for macOS 10.13 (CUDA).
This version supports the new NVidia Ampere architecture on Windows and Linux. Kepler cards with compute model 3.0 are just supported on macOS.
OctaneRender 2020.2 XB2 has superseded this build.
What's new in 2020.2
- RTX always on - and now up to ~1.6x faster!
- Fast Spectral Nested Dielectrics
- ACEScg and filmic processing
- OpenColorIO Management, Display and Export (OCIO profile)
- OpenColorIOColor graph nodes; - Look, Profile and View nodes
- Integrated multi-scatter nodes - Vectron Scatter, Volume Scatter and Surface Scatter*
- Built-in mutli-light, fast fog, and wireframe post processing in imager node*
- User generated compositor AOV passes for combining render and light passes in full a compositor node graph *
- Nested Compositor Operator Nodes*
- Built-in Vectron Primitives and Operators*
- Built-in Raytype, Utility, RGB channel, math and color operator nodes*
- Chaos Texture Tile System*
- UV Distortion Map Textures*
- Gradient Node Improvements*
(*Coming in 2020.2 XB2+ builds)
Roadmap of features now in active development:
- Major Stability (OOP) and out of core memory improvements*
- Billions of unique primitives supported in VRAM*
- New Custom AOV System with LPEs*
- 2X Faster Light Sampling*
- Boolean Geometry And Clipping Nodes*
- Native USD Support*
- GGX mutliscatter and other BRDF improvements*
- Curvature texture node*
- RTX speed improvements*
- Brigade, EmbergenFX and Hydra (Arnold, AnimeRender, Cycles, Storm) Modules*
(*shipping after 2020.2)
RTX always on - and now up to ~1.6x faster!
Octane can now leverage RTX hardware acceleration to evaluate previously CUDA only shaders, such as the new dirt shader.
After removing this limitation we have experienced significant 1.6x+ performance speedups in many production scenes that previously had almost no speed with RTX on.
With this and other RTX improvements, Octane 2020.2 and later will default to RTX on.
Please bear in mind the performance results displayed for 2020.2 XB1 are not final yet since final optimization work is still in progress.
Fast Spectral Nested Dielectrics
We introduce a new feature in Nested Dielectrics to keep track of the interface where a ray is in. This allows for accurate simulation of IOR and improves realism of specular reflections within an interface that has an IOR higher than
Nested Dielectrics brings is the simulation of air bubbles inside another medium. With Nested Dielectrics being able to track the IOR along the ray path, it maintains a list of known IORs which can be used to calculate the interface changes between surfaces more accurately. In this example, where previous Octane versions would have ignored the air bubbles with 1.0 IOR, 2020.2 would find the total internal reflection bounces to reflect off the air bubbles as we try to go from 1.3 IOR (liquid) to 1.0 IOR (air bubble).
IOR behaves differently due to IOR tracking along a ray, notice that the fresnel calculation now takes into account the IOR of the ray path, and this changes the specular reflection of specularly reflective materials within a medium with IOR greater than 1.0.
Additionally, to resolve the ambiguity in overlapping surfaces, here we introduce the surface priority control to allow artists to control the order of preference for surfaces. The lower number suggests a higher priority for the surface material, and a higher priority is preferred over a lower
priority if a ray enters a higher priority surface and then intersects a lower priority surface while inside the higher priority surface medium.
OpenColorIO color management
We have integrated OpenColorIO (OCIO) for render output. After setting up an OCIO configuration, you can select a view and a look to use for the render viewport, or a color space and a look to use when exporting images.
The default ACES OCIO configs can be downloaded from here.
Here's an example of the different looks that can be applied to an image using the Filmic Blender OCIO config:
In addition to OCIO support, ACEScg is now available as a built-in color space when exporting single frames and using batch rendering.
Support for importing image textures in different color spaces (including OCIO color spaces) is planned for an upcoming release.
Integrated multi-scatter tools - Vectron Scatter, Volume Scatter and Surface Scatter
One of the features that makes Octane so powerful is the ability to handle a very large number of instances. Generating instances is usually done via third party tools which generate scatter data that can be imported into Octane.
Octane 2020.2 includes a built-in scatter tool that allows generating arbitrary number of instances according to various patterns that can be configured to achieve many different results.
Example of instance scattering bound by geometry surface:
Another example of instantiation within a volume:
Built-in multi-light, fast fog, wireframe and custom compositing node graph passes
This new system gives artists the ability to change the contribution of individual render passes to the final result without having to start a new render.
This is fully customizable and allows any arbitrary number of inputs. Some common usages would be changing the color and intensity of lights (light mixing), organizing render passes, changing volume color, applying fast fog as a post, overlaying wireframe layers and more.
Preview of new features in active development:
The following is a sneak peek of some of the incoming features that are planned for the release following OctaneRender 2020.2:
Boolean Geometry And Clipping Nodes
A render boolean allows trimming geometry within a volume in the space. This opens a number of new possibilities since it allows effectively modifying the original geometry in real time.
Here is a video of an early prototype showing some of what you would be able to do:
New Custom AOV System with LPEs
The basic idea is to allow arbitrary textures of the scene to be rendered into their own render pass. Most of the time, users use those AOVs to create masks to isolate specific parts of a rendered image. Whenever a texture with a custom AOV pass gets hit by a camera ray, it's value will be splatted into the corresponding pass.
Major Stability (OOP) and out of core memory improvements
Scene management in has been improved over the years yet there are some changes we want to introduce to make the system more robust and resilient to failures:
The original update system stems from a time when Octane as no support for out-of-core and peer-to-peer memory and everything had to fit into memory of each enabled GPU. This may cause problems if data needs to move from one memory pool (local GPU memory or peer-to-peer memory or out-of-core memory).
The changes we intend to do involve the introduction of a new memory management system that will deal with resources differently and deal with errors during the resource allocation as gracefully as possible.
Billions of unique primitives supported in VRAM
Octane will support an unbound amount of geometry per mesh as long as it fits in the available memory.
Other improvements should reduce memory usage such as sharing data between adjacent geometry primitives.
2X Faster Light Sampling
When AI light was first introduced it provided a significant optimization when sampling complex scenes containing both large numbers of images or with image sources that are partially occluded.
In 2021 we intend to introduce improvements to our algorithm which aim is to reduce shading variance even more when a the source of light are mesh emitters, allowing to reduce noise at low sample count compared to both non-AI light sampling and previous versions of AI-light.
Below are equal-sample comparisons of renders using the old an new methods:
Native USD Support
Pixar's USD scene format is getting universal attention due to its unique features and ability to handle large scene with ease, compared to any other scene description format currently available.
We are adding ability to load the USD files into octane. Later our goal is to make the USD format as default format for scene interchange-ability between a plugin and the standalone.
Curvature texture node
Curvature can be useful when creating shaders that take into account how exposed or occluded a given surface point is.
This introduces new artistic possibilities such as worn out surfaces, filling in crevices and more.
Improvements in RTX speed
The new Ampere architecture introduces accelerated hardware support for geometry motion blur.
We intend to introduce support for RTX motion blur and we expect that by leveraging this we can achieve similar performance improvements in scenes that use this feature as we have experienced when RTX support was first introduced for regular meshes.
RTX Support for all primitive types
Even though RTX hardware supports primitive intersection via hardware only for regular triangle meshes, acceleration structure traversal can be also leveraged for other primitive types.
2020.2 XB2 has superseded this build.
Downloads for Enterprise subscription users:
OctaneRender Enterprise 2020.2 XB1 Standalone for Windows (installer)
OctaneRender Enterprise 2020.2 XB1 Standalone for Windows (ZIP archive)
OctaneRender Enterprise 2020.2 XB1 Standalone for Linux
OctaneRender Enterprise 2020.2 XB1 Standalone for macOS (CUDA)
Downloads for Studio subscription users:
OctaneRender Studio 2020.2 XB1 Standalone for Windows (installer)
OctaneRender Studio 2020.2 XB1 Standalone for Windows (ZIP archive)
OctaneRender Studio 2020.2 XB1 Standalone for Linux
OctaneRender Studio 2020.2 XB1 Standalone for macOS (CUDA)
Your OTOY NZ team