2024-04-01

Owned by Carol Payne
Apr 01, 2024
2 min read

April 1, 2024

Host: Carol Payne

Secretary: Carol Payne

Attendees:

Rémi Achard (TSC) - DNEG
Mark Boorer (TSC) - Industrial Light & Magic
Mei Chu (TSC) - Sony Pictures Imageworks
Sean Cooper (TSC ACES TAC Rep) - ARRI
Michael Dolan (TSC) - Epic Games
Patrick Hodoul (TSC) - Autodesk
Zach Lewis (TSC) - Method
Thomas Mansencal (TSC) - Weta FX
Carol Payne (TSC Chair) - Netflix
Mark Titchener (TSC) - Foundry
Carl Rand (TSC) - Weta Digital
Doug Walker (TSC Chief Architect) - Autodesk
Kevin Wheatley (TSC) - Framestore
Mark Reid - Animal Logic

OCIO TSC Meeting Notes

  • <Insert April Fools Day prank here!>

  • ASWF Color Interop Forum recap

    • Close to 30 people attended

    • Link to recording/notes/trancripts in the ASWF slack in the #colorinteropforum channel

    • Mostly introductory

    • First project would be to tackle the "standard" list of colorspaces for use in textures and rendering (to be used in the nanoColor effort specifically, but possibly elsewhere)

    • Spectral was brought up, something for OCIO to maybe start to consider, bit of a chicken & egg situation

    • Remi: can we get someone from ICC at the forum? They're already doing spectral things, we should be at least communicating with them (Carol will reach out) 

  • Opinions on PRs that clean up whitespace

    • Remi: whitespace should be in separate commits using the .gitignore 

    • Would that work with squash and merge?

    • Not sure, we should test

    • Remi also looking into clang format stuff, should we wait for that?

      • Some things will definitely need exceptions, but should mostly work

      • Timing wise: when should we do this?

      • Michael: should be isolated to it's own thing so it's easy to track. Larry did this with OIIO - Remi followed their example. Can ask Larry how he handled this timing wise 

      • Mark R - also messes up backports to earlier versions

    • Should use ruff? for python

  • ACES 2 updates?

    • Remi: still doing minor updates, still working on the blues rendering issue. Might be addressed as an LMT. Scott still working on the CTL, should be ready in the next few weeks. 

    • Doug - is the inverse working in the python?

    • Remi - yes, the two step is mostly for HDR is seems, so the. numbers are worst there. Will try and send some things out to the group to review. 

    • Remi - need to discuss at some point the amount of parameterization we want to provide via OCIO via FixedFunctions

    • Also need to talk about the initialization phase for the pre-cal LUTs - Doug thinks we can look at what is already in the inverse LUT functions in OCIO for a start.

  • Minizip build issues

    • Not clear that minizip is going to change on their side. But they do have an ASCII version we could switch to instead, need to dig a bit further

    • Open PR to switch OCIO likely too simple, should wait on merging it. Would likely need a test to check minizip version at least. 

    • Also with the Conan PR - Remi not sure it's exactly what we want. It's a bit too brute-force as is. There's a thread on the #dev channel talking about it. 

  • Review “To Close” Issues

    • We'll go through this list at the next TSC, if you have some time to take a look before that, would be much appreciated.  

    • Mei: should we have an automated bot that closes an issue after a certain time?

    • Thomas: having a human do this step is ideal, these auto bots can be annoying and too agressive

    • Carol: yeah, we will likely lean into the automation features in GitHub projects for organization, but leave the actual issue status updates to humans for now, until we get a good process in place that we like, at least.