Color Checker and Log Footage - Bug?

Dear all,
I am trying to color-correct footage that has been recorded in Log (HLG). I am using a Color Checker by XRite (Video Passport).

As a node setup, I am using the following:

Node setup

Bildschirmfoto 2021-10-12 um 21.58.38.png (125.34 KiB) Viewed 1535 times

The first node makes a color space transform - from Rec2020 HLG to Rec709 Gamma 2.4 (and a bit of level adjusting using the primary wheels, but that seems not to make a "real" difference concerning the problem). In the second node, I use the color match function, which automatically reads the values out of the color checker. Here, I can chose the input gamma as well as target gamma and color space.

In my opinion, as I was using a color space transform in the first node, I should use a input color gamma related to Rec709. However, this gives me this output:

Rec709_Image

Bildschirmfoto 2021-10-12 um 21.56.28.png (563.34 KiB) Viewed 1535 times

As you can see, the values are way off! If using an HLG input gamma, the result looks much better:

Rec2020_Image

Bildschirmfoto 2021-10-12 um 21.57.28.png (520.63 KiB) Viewed 1535 times

So to my question: Shouldn't the color match function read the gamma which comes out of the color space transform node? If not: How do I correctly apply the color match function? Is it ok to leave out the color space transform and directly use color match with the correct gamma?

Thank you very much, best
Christian

I am interested as well. I have an Canon r5 and tried a few different methods and was never happy with the outcomes.

I settled on not using automated tool at all.

Second node, CST of Clog 3 to Davinchi Wide gamut (DWG)

Third node, white balance

Fourth, primaries and everything else.

6th node CST to DWG to 709


First node. I use the Hue-Hue and Hue-Sat curves to manually adjust so each chip lines up on the vector scope. See this guide at 6:17

Yeah, watched many tutorials and read in the new color guide by BM, but couldn't find an advice on this topic.

Unfortunately, your video does not load - however, I know this process using the vectorscope.

Can you tell me
- why you are using a conversion to DaVinci Wide Gamut? If your timeline is (and I assume it is) Rec709? If you are using correctly tagged files and activate the internal DaVinci Color Management Workflow, shouldn't the conversion (which we try to rebuild using CST) also do a conversion to Rec709?
- If using a color managed workflow, the (automated because of the file tags) CST comes first - all other corrections after that. So wouldn't it be logical to do your Hue-Hue and Hue-Sat corrections as the second node?

In the end, this doesn't solve my issue with the color match feature, but perhaps gets new insights

Best
Christian

If you are using correctly tagged files and activate the internal DaVinci Color Management Workflow, shouldn't the conversion (which we try to rebuild using CST) also do a conversion to Rec709?

I do not use color managed workflow, it is completely broken with Fusion and BM has not yet even acknowledged how bad it is, so it is unlikely they will fix it soon.

viewtopic.php?f=22&t=137485&p=745551&hilit=wide+Gamut#p745551


The crazy work arounds for the problem are way above me, so I use CSTs in color tab. I don't use 709 as the HDR tools don't work right with 709.

I am also not using a color managed workflow as my files don't contain the correct metadata.

To come back to my first question: Do you know what "Source gamma" of the Color Match Function means? Is it the Gamma of my source material or the Gamma of what comes out of the previous node? So if I place a CST node before Color Match, it is more logical to me to use the output gamma of the CST as source Gamma of Color Match, but strangely, this gives strange results. Using the source Gamma as recorded by the camera seems to work, but I don't understand why...

Thank you, best
Christian