ACES 2.0 is the second major release of the components that make up the ACES system. The most significant change is a new suite of rendering transforms whose design was informed by collected feedback and requests from users of ACES 1. The changes aim to improve the appearance of perceived artifacts and to complete previously unfinished components of the system, resulting in a more complete, robust, and consistent product.
Highlights of the key changes in ACES 2.0 are as follows:
New output transforms, including:
A less aggressive tone scale
More intuitive controls to create custom outputs to non-standard displays
Robust gamut mapping to improve perceptual uniformity
Improved performance of the inverse transforms
Enhanced AMF specification
An updated specification for ACES Transform IDs
OpenEXR compression recommendations
Enhanced tools for generating Input Transforms and recommended procedures for characterizing prosumer cameras
Look Transform Library
Expanded documentation
Rendering Transform
The most substantial change in ACES 2.0 is a complete redesign of the rendering transform.
ACES 2.0 was built as a unified system, rather than through piecemeal additions. Different deliverable outputs “match” better and making outputs to display setups other than the provided presets is intended to be user-driven. The rendering transforms are less likely to produce undesirable artifacts “out of the box”, which means less time can be spent fixing problematic images and more time making pictures look the way you want.
Key design goals
Improve consistency of tone scale and provide an easy to use parameter to allow for outputs between preset dynamic ranges
Minimize hue skews across exposure range in a region of same hue
Unify for structural consistency across transform type
Easy to use parameters to create outputs other than the presets
Robust gamut mapping to improve harsh clipping artifacts
Fill extents of output code value cube (where appropriate and expected)
Invertible – not necessarily reversible, but Output > ACES > Output round-trip should be possible
Accomplish all of the above while maintaining an acceptable “out-of-the box” rendering
what are 3D Gaussians? They are a generalization of 1D Gaussians (the bell curve) to 3D. Essentially they are ellipsoids in 3D space, with a center, a scale, a rotation, and “softened edges”.
Each 3D Gaussian is optimized along with a (viewdependant) color and opacity. When blended together, here’s the visualization of the full model, rendered from ANY angle. As you can see, 3D Gaussian Splatting captures extremely well the fuzzy and soft nature of the plush toy, something that photogrammetry-based methods struggle to do.
Spectral sensitivity of eye is influenced by light intensity. And the light intensity determines the level of activity of cones cell and rod cell. This is the main characteristic of human vision. Sensitivity to individual colors, in other words, wavelengths of the light spectrum, is explained by the RGB (red-green-blue) theory. This theory assumed that there are three kinds of cones. It’s selectively sensitive to red (700-630 nm), green (560-500 nm), and blue (490-450 nm) light. And their mutual interaction allow to perceive all colors of the spectrum.
“Simon Willison created a Datasette browser to explore WebVid-10M, one of the two datasets used to train the video generation model, and quickly learned that all 10.7 million video clips were scraped from Shutterstock, watermarks and all.”
“In addition to the Shutterstock clips, Meta also used 10 million video clips from this 100M video dataset from Microsoft Research Asia. It’s not mentioned on their GitHub, but if you dig into the paper, you learn that every clip came from over 3 million YouTube videos.”
“It’s become standard practice for technology companies working with AI to commercially use datasets and models collected and trained by non-commercial research entities like universities or non-profits.”
“Like with the artists, photographers, and other creators found in the 2.3 billion images that trained Stable Diffusion, I can’t help but wonder how the creators of those 3 million YouTube videos feel about Meta using their work to train their new model.”
