2024-03-04
March 4, 2024
Host: Carol Payne
Secretary: Carol Payne
Attendees:
OCIO TSC Meeting Notes
GitHub Actions Updates
Remi - his fixes should be good for github actions
PR ready to merge
Analysis workflow - github secret is at the ASWF level, working on that fix
SLP 2024
Learner application is open, please help us spread the word!
Doug - worth reaching out to SMPTE
ASWF Color Interop Forum announced - March 25 @ 12pm PT
NanoColor Working Group Charter Review
Kevin - not super clear with the charter what the deliverable is - is it code? is it standards?
Do we need to be a little clearer about the fact that it is code that is being deliverable
Clearer that it is an implementation and not just an API
Jonathan adding language to clarify
Mark B - saying header only might be cart before the horse
Mark B - document might be missing why this is necessary in the first place - what issue are we trying to solve?
Nick Porcino - point 2 is maybe the key point here - the need for a stable, immutable set of standard color spaces
Point 3 is also important - analytic closed form equations for portability, archivability, etc
Doug - also minimizing dependencies from the main OCIO library
Jonathan - also important for MaterialX to be able to make this version of OCIO a required dependency, not optional. Color information is not optional for materials, it is essential to understand intent.
Jonathan - other applications such as three.js and gLTF
Mark B - concerned about a growing list of immutable colorspaces - one of the main tenants of OCIO is flexibility for the author to define their color pipeline
Kevin - does this include standard display colorspaces?
Doug - not for the MaterialX / USD use cases for nanoColor, but yes for OCIO Lite for use in a web browser. Would include some sets of analytically defined, closed form, invertible transforms including some display / view transforms
Kevin - what about spectral?
Doug - very interesting questions, but as none of these systems currently support spectral data, out of scope for the original project
Mark - what is the big difference between these libraries?
Doug - list of ops would not include LUT 1D, LUT 3D or file transform. Also the dependencies as you mention, zlib, yaml, expat.
Thomas - where is the delineation? Will main OCIO depend on this new library?
Doug - this is one of the main questions the group will need to figure out
Michael - ideal would be for main OCIO to be a client or dependent on this new lite/nano library
Nick - this would also be ideal for USD
Kevin - what about white point adaptation etc? Materials aren't like our eyes
Doug - yes, white point adaptation cats would be defined as part of the standards
Mark - what about float vs half float?
Nick - yes, working in float
Sean - Matte painting workflows are definitely something to consider
Carol - there will be much more to discuss and work towards as we start to think about how this implementation will actually work. Stay tuned for working group meetings and things once we talk with the TAC on Wednesday, and feel free to slack / comment on the doc in the meantime.