I had my monitor calibrated this afternoon and with normal rec709 out of Resolve we hit our colours bang on. When we changed the Resolve to a Aces cct setup IDT ADX 16 and ODT Rec709 we were getting reading that were off in yellow and cyan.
Any ideas what could be the issue?
I’m not 100% sure I understand your test but it’s worth noting ACES is very different than a legacy rec709 workflow. Passing a pure ACES color (e.g. ACES yellow [1, 1, 0] ) through the rec709 ODT isn’t expected to yield a rec709 code value of that color (e.g. Rec708 yellow [1, 1, 0]). You can use the inverse rec709 ODT to find the exact ACES values required to produce a particular display code value.
I’m grading in
Rec709 My monitor is calibrated correctly with RGBCMY all correct. When I swop over to ACES RGBM are correct. C and Y don’t sit where they should with a probe. So I can not trust what I am seeing on screen when grading.
are you using 16 BIT DPX scans as source footage?
How exactly are R, G, B, C, M, and Y being generated? In what space?
Yes my source files are 16bit log DPX files from a Neg scanned on Scanity.
I have film neg scans. My Resolve is setup in Aces cct with an ADX IDT and ODT of Rec709. When calibrating the monitor the yellow and cyan are incorrect according to a probe and on the vectorscope. My reference generated as an offline clip which is a rec709 QT has the rec709 IDT. This is showing the yellow and cyan correctly. My problem is I cannot accurately match to the reference which is critcal for restoration.
When you say “the yellow and cyan” are you talking about colour bars? As has been mentioned elsewhere, there are no ACES values which will produce output through the Rec.709 ODT which hits all six colour bar targets. But less saturated colours can pass through the Rec.709 IDT and Rec.709 ODT unaltered. If your reference contains image colours as saturated as the bars, you will not be able to reproduce them with ACES.
Yes the colour bars but also the picture. For example shots in a church with stained glass windows the yellow especially was desat. So what you’re saying is this will never match in ACES?
Never is the wrong word – the inversion problem and desat yellow is high on the list to fix for ACESnext.
Great thanks for all the help.