July 7th, 2021

Host: Doug Walker

Secretary: Carol Payne

Attendees: 

  • Doug Walker (TSC Chief Architect) - Autodesk
  • Carol Payne (TSC) - Netflix
  • Kevin Wheatley (TSC) - Framestore
  • Thomas Mansencal - Weta Digital
  • Zach Lewis - Method
  • Nick Shaw - Antler Post
  • Chris Brejon - Illumination

Apologies:

  • Michael Dolan

OCIO Config Working Group Meeting Notes

  • CG Config:
    • WIP Code here: https://github.com/AcademySoftwareFoundation/OpenColorIO-Config-ACES/tree/feature/cg
    • Test artifact/config generation: https://drive.google.com/drive/folders/1TcbM9_xy9e3KqwKz78SFsC7VvgnnkLH-?usp=sharing
    • What should the ID tag in the generated CLF files be? Sort of an ACES TransformID, even though they aren't official ACES transforms? We'd like to have a way to reference it between configs. 
    • Some discussion relative to having a read method for GroupTransform, Doug and Thomas following up about the current way to achieve this in python. 
    • Studio config should just "work" once the CLF bits are worked out, and some details around naming and other minor functionality details are done for the CG config.
    • Should we follow the naming convention for the built in transforms? Seems to make sense, good for consistency.
    • So the two big to-do's are the transform IDs and the CLF names. Will add columns to the spreadsheet so we can start to suggest naming conventions.
    • Doug: first part is python to generate CLF's with specific transform IDs. Second part is treating the CLF's as CTL and parsing them to generate a config that points to the CTL file. 
    • Zach: should OCIO provide some sort of configurable colorspace object? Zach will provide a code example to demonstrate what he's asking to the group.
    • Kevin: is it worth trying to parse and "re-bake" the transforms into the OCIO config where possible based on the CLF? To eliminate dependency on external files? 
    • Is this the role of the config generator or the role of OCIO itself? Having the ability to read in a config FileTransform or CLF and give me back a Group Transform for it, if possible. 
    • Thomas: we could technically do this now.
    • Doug: this does exist, you can take a file transform, and make a processor based on it, and it will give you the group transform. 
    • Group does agree that limiting dependencies on external file transforms is a good goal.
  • Camera Vendor Outreach:
    • Two parts: Outreach to vendors that already have ACES IDTs, and to vendors that don't have official IDTs around inclusion in the studio config.
    • Collaboration to be done with the ACES IDT Working Group, as well as the AMF working group around IDT implementation and consistency between applications. 
  • OCIO 2.1 / ACES Gamut Compression:
    • OCIO 2.1 coming soon
    • Thomas: have updated the reference config to include the Reference Gamut Compression, it's in a draft PR, Thomas will push that through and add a gist of what it generates. 
    • It's a look transform with ACES 2065-1 in and out. 
    • This is fine for the reference config, but might need more of a discussion for the studio config etc as far as end user. 
    • A lot of discussion around the ability to test these new features in OCIO 2.1 / the reference config once created. In the end, we'll likely need folks to build the source and use the command line tools to test until integration with DCCs catches up.
  • Items for next meeting agenda:
    • Discuss CG Config Progress, talk about where folks can jump in to help


  • No labels