CMY Splitter/combiner nodes?

QUESTION: Can Davinci Resolve do CMY splitter/combiner nodes?

Backstory: I'm processing 8mm Kodachrome home movies from the 40s to the 60s (I have a bunch of Super-8 Kodachrome-II (and some Ektachrome-160 that I shot in the 70s thru the 90s that I also want to get scanned and graded).

The recent scans are done on a Lasergraphic ScanStation at 16-bits per pixel color, so I have a lot of 'meat' to grade with.

With some of the shots I've graded, the different dye layers have faded at different rates, and are consequently hard to grade with one main node and the color-pucks. I had much better luck with the RGB splitter/combiners, and a node for R, G, and B. but it still takes some fiddling with the slope of the curves and tweaking of the exposures in the HDR grader, before doing final adjustments in the node after the combiner. so when I tweak Red, I'm actually tweaking both the Yellow AND Magenta layers, Green is Cyan+Magenta, and Blue is Cyan+Yellow.

Ungraded shot

Faded_1.245.1.jpg (250.05 KiB) Viewed 791 times

It's difficult to grade the above image.
This grade is after tweaking of curves in the RGB splitter/combiner node rig. I concede that maybe this is the best the image will ever be, as it was also overexposed when shot.

My point is, I suspect that the tweaking would be easier and more intuitive if I could approach it as which die layer faded, and control that, rather than which dye layer faded, and how do I manipulate the R, G and B curves to attempt to correct for the one faded layer without also affecting the other layers.

So is there a way to split/combine CMY rather than RGB do that (yet)?

For the future, I'd like to have the capability to tweak Kodachrome dyes or Ektachrome-reversal dyes, with splitter/combiners tweaked to the spectral characteristics of those chemistries.

I do archival film work all the time and -- to me -- you're just as well off using secondaries to do what you describe. One thing that's tricky is when you have density flicker issues that are different in the Red, Green, and Blue channels. You can make a case for using a splitter-combiner node and attacking the flicker that way.

Shading problems are another issue, where one side of the film image is a slightly different color (or density) than the other, and I tend to use power windows -- like a straight vertical window on one side of the frame -- to fix problems like that.

I don't see a need to split or filter different film dye layers separately from the other, because to me, everything affects everything else, particularly with different luminance levels. Again, the flicker/color "breathing" problems with old Kodachrome are going to be a bigger issue, as well as fading color. Overall density, particularly underexposed home movies, are a nightmare to deal with.

As for color, the yellow and cyan layers tend to fade fastest in print film, which is the reason why a lot of old film disintegrates into magenta. Once those layers are gone, it's hard to reconstruct normal color by any means, though you can rebalance whites into white and shadows to black to some degree.

Okay, I need to study up on DR's methods/tools for controlling the 'secondary' colors. The 3,000+ -page manual is a pain to use though — any search takes the computer 5 to 10 minutes to find the pages containing the word searched for.

the yellow and cyan layers tend to fade fastest in print film, which is the reason why a lot of old film disintegrates into magenta

Boy howdy

As I mentioned. I have some Ektachrome-160 reversal that scares me for that reason. So far, the C&Y layers are holding up (knock on wood), but those reels are the first to be scanned when finances allow...

As to the Kodachrome, those dyes resist fading for a lot longer, with the caveat that the material be kept in darkness when not being used. It seems like exposure to light is what breaks down the Kodachrome dyes...

One thing that's tricky is when you have density flicker issues

Lucky for me, not-so-much with the footage I'm working with. Also I'm not making it perfect, just as good as it reasonably can be made.

Shading problems are another issue, where one side of the film image is a slightly different color (or density) than the other, and I tend to use power windows -- like a straight vertical window on one side of the frame -- to fix problems like that.

I've done exactly that. The footage I'm working with has a red cast around the sprocket holes, probably because of the extra agitation the sprocket holes create when running through the developer machinery (and Kodachrome processing was extremely long, complex, and toxic.) I use a node with a power window before the main node on the Red channel. Again, knocking on wood, I haven't seen the same deviation on the other layers, but