Difficult to fine tune colors in Resolve

[Toby Tomkins]

Sounds like you could be a Tetrachromat!

But to answer your question, use a node’s output gain to fade the adjustment node effect from 0-100%

[Juan Salvo]

Can you describe a bit more about how you are working? Are you adjusting under a lut, in aces? Do you have a node that’s increasing your saturation perhaps? How are you monitoring?

https://JuanSalvo.com
https://theColourSpace.com

[Itai Bachar]

Im “LUTing” it manually, just adjusting some primaries and curves,
nothing out of the ordinary, monitoring on a calibrated apple cinema.
I’m attaching a screenshot, showing the greenish cast achieved by 0.01 gamma change. it is really a struggle currently to fine tune this.
(I am new to resolve)
Thanks

[Kevin Cannon]

Hi Itai, could you show your node tree and give your order of operations?

Keep in mind that if you put the LUT (or corrections that do the same) in a later node (say node 3), that will drastically exaggerate changes you make in nodes 1 and 2. But if you add a node 4 the corrections you make there won’t be multiplied by the LUT, so to speak. For that reason I like to set up three node: 1 is for rough adjustments to the log image, 2 is for the LUT (or corrections that do the same) and 3 is for fine adjustments.

KC

Prehistoric Digital

[Joseph Owens]

[Itai Bachar] “nothing out of the ordinary, monitoring on a calibrated apple cinema.”

For starters, you’re more than likely monitoring in 8-bit, so what you think is a .001 change is quite a bit more than that even if you are working in full scale 0-255, less if you’re working in 709. Hues are going to pop in large increments, more than one degree at a time, if you do the math. (360/255, roughly, at full saturation.)

If not for any other reason, the coarseness of an 8-bit display, which describes most graphics monitors, disqualifies them for grading. Apple Cinema Displays are particularly bad, “calibrated” or not.

As far as tetrachromatic color vision goes, I thought that was genetically impossible for males, my understanding being that the trait resides on the second X-chromosome. But that reflects an imperfect understanding on my part, probably.

jPo

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Itai Bachar]

Hi Kevin, my node tree is actually similar to what you are describing,
I do the LUT (either Arri lut from the 3d lut menu, or my own “lut”),
than minor tweaks.
So I guess no un-necessary “multiplications” happening,
have I got it right?
Thanks

[Itai Bachar]

Hey Joseph
Thanks for the scientific explanation, I guess I’ll need a pro grade monitor one of these days.
In what way apple monitors are worse?
Can you recommend a monitor (a computer monitor) that will be better for the job than apple cinema, or they are all un-qualified?
What about those Eizos and NEC pa’s, or any other high end pc monitor?
Or is that will always be a compromise and I need a pro broadcast monitor?
those work in 10 bit?
Thanks

Tetrachromatic? you guys can see the green cast in the screenshot above,
no need for special female genes… just plain eyesight…
🙂

[Toby Tomkins]

That is an oddly large change, 8 bit or not! The gene has only been discovered in women but I don’t think there’s a limit on the genetics side to women but I might be wrong, it was just a throwaway comment as a joke! (-;

I think something is wrong with your install or set up. Try reinstalling the latest version maybe? If you do t want to do that then use my workaround described in the first reply I gave! This will also test the 8 bit monitor issue but I highly doubt this is the case.

[Itai Bachar]

Yes I agree this is odd, though I did encounter this “narrow color space”
before in Resolve, but I have just a few projects under my belt with it.
Your workaround is useful, and your joke is useful too! 🙂
thanks

[Marc Wielage]

[Itai Bachar] “Hi Kevin, my node tree is actually similar to what you are describing, I do the LUT (either Arri lut from the 3d lut menu, or my own “lut”), than minor tweaks.”
What happens if you eliminate the LUT entirely and just use the first node to simulate the actions of the LUT? I have encountered Log-C situations before where I just went at it with no LUT and did just fine.

Bear in mind that in The Old Days, we color-corrected for decades without LUTs using film, which is always photographed Raw, essentially in Log space. It can be done, it’s just a question of where you need to go and what you’re trying to do.

BTW, I always find it’s best to use the first node to get the image to something approaching normal, then do all the fine-tuning downstream from there. Whether I use the LUT or not is dictated by the material.