Recap:
- @carolalynn pointed out this ACES Central post as info from the Architecture group around including the compression in the AP0 > AP1 transform vs. as a standalone operation:
-
Notice of Meeting - ACES Gamut Mapping VWG - Meeting #15 - 6/4/2020 - #15 by carolalynn
- Things in there include a few workflow pipelines from @Thomas_Mansencal, which illustrate the complications not just for application/tracking of the gamut compression itself, but also versioning and tracking of the ACES system overall.
- Discussion seemed to end with support for sticking with the standalone LMT-like operation.
- Parameterization Discussion:
- ACES version vs Non-ACES with parameterization
- @mario noted that just removing “ACES” from the name might not be enough - the two should have separate and distinct names.
- @carolalynn: Two ACES versions would be challenging, as technically with the CTL you can see the constants and edit them if needed. It’s up to the applications on how they expose or don’t expose those parameters in a non-ACES, creative version.
- @michaelch: Would be useful to have a standard for the operation across applications. Also agree with Mario on separating the tech “fix” vs. the creative operation
- @mario: I want to avoid talking about parameters with clients. In ACES, if you say it’s gamut compressed, everyone should know what we mean, there should be no confusion. The creative should be an artist tool, just like other things you do in comp.
- @matthias.scharfenber - and then how do we store those parameterized values? Yet another sidecar file? Tracked via AMF, or in a container header, or both?
- @jzp: an interesting question is around deliverables: what happens when a studio wants a linear AP0 delivery? Presumably the shots from VFX would have the gamut compression applied, but the non-VFX shots would not. Sort of already living that with the current blue light LMT. A note for the workflow documentation.
- @jzp also asked about 3D LUT application, as ACEScct is not acceptable. We talked about the possibility of a custom shaper as referenced in the Architecture Documentation, or starting from a vendor encoding gamut (ala @nick’s camera magic)
- @carolalynn : if the agreement is that we should definitely have the immutable version and a separate parameterized version (whatever that looks like), I propose that this group focus first on the static ACES version in applications first, and then learn from that, and revisit parameterization at a later time. We will revisit this proposal with the EMEA group next week.