Thursday, September 17, 2020
5:00pm - 6:00pm Pacific Time (UTC-12:00 Friday)
Please join us for the next meeting of this virtual working group (VWG). Future meeting dates for this month include:
9/24/20, 9:30am
Dropbox Paper link for this group:
We will be using the same GoToMeeting url and phone numbers as in previous groups.
You may join via computer/smartphone (preferred) which will allow you to see any presentations or documents that are shared or you can join using a telephone which will be an audio only experience.
Please note that meetings are recorded and transcribed and open to the public. By participating you are agreeing to the ACESCentral Virtual Working Group Participation Guidelines
Audio Only
You can also dial in using your phone.
Dial the closest number to your location and then follow the prompts to enter the access code. United States: +1 (669) 224-3319 Access Code: 241-798-885
It’s rare that a production is willing to share content, especially content that was ‘difficult’, so I want to publicly thank Andrew Savinsky, a compositor on the Ukranian TV series ‘Sniffer’ who contacted me about the sequence below several years ago, for getting permission from Pronto Films, the content owner, to share this content for “demonstration purposes”.
The image below is threefold:
on the top, I’ve grabbed the ALEXA K1S1 rendering;
in the middle (grabbed from Nuke) the ACES P3D65 rendering without gamut mapping, and
on the bottom, the group’s gamut mapping algorithm with the default parameters.
I’m not sure images here appear at full resolution, but three full-resolution plates from this sequence have been uploaded to the group’s dropbox.
@LarsB notes he’s already had to push the power coefficient below 1 - down to 0.5
@LarsB also notes that the algorithm causes noise crosstalk between channels, and can increase appearance (typically from blue to green). Suggested synthetic image to test.
@joseph recommends shooting an over/under for a test, also deliberately mess up the color balance in raw
@sdyer is uploading the Garlic film scan for us to take a look at to double check efficacy with film applications
@carolalynn stated a proposal for how we might run user testing, will surface this to the other time zone and add things to the dropbox paper with more info.
@carolalynn will also be sending a general “update” mail/post to ACES central and the original group list to elicit more feedback and source candidates for testing
@matthias.scharfenber showed his Nuke script with a modified shaper space to accommodate large negative values to then encapsulate the algorithm into a 3D LUT. We are not sure of the benefits of putting this out there, as it’s not the ideal implementation. But could be useful for OCIO 1.0 applications possibly, with many cautions and red flags and warning signs.
Wanted to follow up on this. Is there an available OCIOv1 config available (with red flags)? Or maybe a code snippet?
Also I wanted to make sure I’m getting the right/current shaper. I can download the Nuke script from @matthias.scharfenber and make the shaper with it. Is that the best version to use?
The plan, at least on my end is to NOT produce an official OCIO v1 Config nor support ACES 1.3 via the Legacy Repository, we are trying to archive it and let it rest-in-peace.
Understood. So as far as OCIO goes, it would be updated here in OCIOv2, correct? That would not be a 3D LUT, but the proper implementation of the gamut compression based on CTL. Definitely looking forward to that!
What I was asking about here is in regards to legacy software using a 3D LUT and shaper. Just want to make sure I’m getting things from the right place, and not mucking it up
Legacy software might be kind of a misnomer here .
If I understand correctly, gamut compression needs to be a Fixed Function and thus a Builtin transform in OCIO. The gamut compression builtin transform will be included in OCIO v2.1. Software would need to update to support OCIO 2.1.
