Please Check My Workflow - Colour Management

Hi team,

Although i do currently work roundtrip with FCP im really struggling to achieve consistency out of DR alone, (irrelevant of monitor). This has been endlessly discussed, but im yet to find an answer that can get my exports remotely resembling what i see within DR.

Currently experiencing a wild lifted gamma shift & less saturation. Trying to counter this shift by bringing exposure down in FilmConvert by -0.3 across all clips prior to export, far from ideal.

Pref
- Color Science > ACEScct
- IDT > Slog3.cine
- ODT > ACEScct

Deliver
- ProRes 4444 XQ
- Data levels > Auto
(Auto/video are very washed out & Full is clipping & crushing)

- ODT > Rec709

Hi Dmitry,

Change the ODT (in color management pref) to 709? This means that the footage is 709 before hitting nodes, ie. filmconvert.

I've also tried various color/gamma tags (709) on export with nothing resembling close to what im monitoring in DR.

Okay, on delivery:
'CST 709 + Gamma Tag 2.4' plays back in quicktime very similar - however when i drag that file into FCP the gamma shifts again. I must be getting warmer here.

I guess it is impossible to use FilmConvert with ACES (as well as many other tools and options). At least you need to understand how color spaces and gamma transforms in ACES under the hood before use it.

I can suggest stay away from ACES until you really need it for some specific work.
Try to learn this more flexible YRGB workflow viewtopic.php?f=21&t=65149#p537852

Also read Explanation about Gamma and Color Shift Problems here:
viewtopic.php?f=21&t=101253
viewtopic.php?f=21&t=114047

Thank you Dmitry - you're correct that i may not need to worry about ACES for the moment.

In an attempt to solidify 'a workflow', I'm currently using this ACES workflow from Netflix: https://partnerhelp.netflixstudios.com/ ... in-Resolve

FilmConvert etc work as expected but i now cant find a CST/gamma tag that works coming back into FCP. As i mentioned above, exporting CST 708 2.4, looks the same in quicktime, however when dragged into FCP, it shifts back to 2.2

- brain exploding..

I would make it a practice (like saving often etc) to manually set gamma interpretation wherever possible. Relying on metadata tagging is notoriously unreliable.

Regarding ACES, project level ODT does not specify what the state of data is before nodes but after. If your project is set as ACEscct then this is what hits the color nodes too. By setting ODT to ACEScct you write cct log to prores and it is pretty far from rec709 as you discovered already, because it is ACEScct log data.

You are using a properly calibrated external monitor connected to a dedicated I/O device, yes? A Decklink or similar?

You are using a properly calibrated external monitor connected to a dedicated I/O device, yes? A Decklink or similar?

Not on this remote setup at the moment. Calibrated MacBook screen only.

The issues I’m having are before monitoring though.

You are using a properly calibrated external monitor connected to a dedicated I/O device, yes? A Decklink or similar?

How exactly does decklink IO change the order of color transforms in Resolve? I'm eager to know because it seems to be a solution to every problem at hand.

I would make it a practice (like saving often etc) to manually set gamma interpretation wherever possible. Relying on metadata tagging is notoriously unreliable.

Regarding ACES, project level ODT does not specify what the state of data is before nodes but after. If your project is set as ACEscct then this is what hits the color nodes too. By setting ODT to ACEScct you write cct log to prores and it is pretty far from rec709 as you discovered already, because it is ACEScct log data.

Thanks for this Hendrik, I’ve read your post a number of times. On my graded project, when I change the ODT to 709, it appears as 709 on top of 709 because of my filmconvert node - so I’m really not sure what the correct process is here.

Clearly still getting my head around not only color management in DR, but the double dipping, round trip to FCP & the complications it’s causing.

I feel I originally tried every gamma setting but couldn’t replicate once back into FCP. I’d honestly love to know if anyone has consistent results here.

As mentioned, my ACES process seemed fine when rendered as CST 709 2.4 - this playback within QuickTime close enough.

However - once this footage lands inside FCP the gamma shifts to what I imagine is 2.2, as it’s identical to rendering the above as CST 709 2.2

However - once this footage lands inside FCP the gamma shifts to what I imagine is 2.2, as it’s identical to rendering the above as CST 709 2.2

It is the interpretation of data (what software thinks is in file) that shifts, not data itself (actual values in file). Force gamma interpretation to 2.4 in FCP and it should look the same. Interpretation has effect only if some kind of color management is active, so if in doubt you can always try comparing your renders in Resolve YRGB noncolormanaged project and see if they match.

However - once this footage lands inside FCP the gamma shifts to what I imagine is 2.2, as it’s identical to rendering the above as CST 709 2.2

It is the interpretation of data (what software thinks is in file) that shifts, not data itself (actual values in file). Force gamma interpretation to 2.4 in FCP and it should look the same. Interpretation has effect only if some kind of color management is active, so if in doubt you can always try comparing your renders in Resolve YRGB noncolormanaged project and see if they match.

Yes, definitely appears to be the interpretations.

I can't seem to find a way to view or change gamma interpretation within FCP.

Just to follow up here in case someone runs into the same issue (which ive found a couple of others did).

Rendering from DR as CST 709 2.2 comes into FCP as 2.2
CST 709 2.6 comes into FCP as 2.6
However CST 709 2.4 comes into FCP as 2.2

My workaround is to change my system color science to 709-A & re grade.

Just to follow up here in case someone runs into the same issue (which ive found a couple of others did).

Rendering from DR as CST 709 2.2 comes into FCP as 2.2
CST 709 2.6 comes into FCP as 2.6
However CST 709 2.4 comes into FCP as 2.2

My workaround is to change my system color science to 709-A & re grade.

viewtopic.php?f=21&t=101253

I guess it is impossible to use FilmConvert with ACES (as well as many other tools and options). At least you need to understand how color spaces and gamma transforms in ACES under the hood before use it.

One big flaw with using traditional LUTs or "looks" like Film Convert is that they will not give predictable results in different color spaces, like HDR. There's a lot of reasons I try to avoid them whenever possible and just use the normal controls within Resolve (with rare exceptions).