10-bit HEVC to DCP workflow
Hi all - I have an extremely specific workflow question, but I'm hoping someone might be able to help me unravel a couple things.
I work for a cinema chain, and my end deliverables are DCPs created using the CuteDCP plug-in in Premiere. My question is about the best workflow to preserve color space when working with 10-bit source material. With increasing frequency, I'm editing footage either shot on a Sony FS7 in the XAVC-I codec at 10-bit 4:2:2, or material from consumer UHD BluRays, transmuxed to 10-bit HEVC mp4's at 4:2:0, which CC2020 plays quite nicely with.
One of my biggest limitations is that we don't have 10-bit monitoring available (believe me, it's on my wish list). So there's a decent amount of educated guessing going on. With FS7 material, I usually apply a LUT to make the slog3 footage look good in an sRGB color space, and call it a day. What I see on the big screen when DCPs (with their XYZ color space) play back through the projector usually looks pretty 1:1 from what I see on my 8-bit monitor. My guess is I'm sacrificing a little chroma and luma depth using this workflow? Although in theory Premiere should be preserving that depth when I drop the footage into a timeline? It's hard to say how much is power of suggestion.
Of course with HEVC material from consumer discs, I only want to transpose the already-color-corrected material as accurately and richly as possible from REC2020 to the DCPs' destination XYZ color space. Naturally with my 8-bit monitor the gamma is transposed very inaccurately, but this is to be expected and I can throw a temporary saturation/contrast effect on top of it for the sake of my eyes and brain as I work. However, if I leave the material untouched, should CuteDCP's color conversion not leave the bit depth intact as it converts to XYZ? (I realize CuteDCP is a very niche product and maybe I should be asking them). I'll definitely do some tests ASAP, but does anyone have any knowledge about how color spaces are preserved and converted through a workflow like this? Would the "maximum bit depth" check box under sequence settings have any impact? I greatly appreciate any help folks could offer.
