Sat Apr 03, 2021 9:22 am
I hope you’re only using this as an example. If so, please excuse the following rant:
Shooting only* red light is a mistake that only makes postproduction more difficult. I’ve seen a few DPs do it just because they have access to some kind of colored LED... astra, Helios, whatever... It drives me nuts. You’re throwing away 2/3 of the information that’s hitting your sensor. Stop. You would literally be better off filming in black and white. If you want the image to be red in the end, it’s much easier to do *that* in post, and preserve the other channels for things like tracking, keying, or whatever.
In fact, the easiest way to get the effect you’re after would be to film an actress with blue lenses in full color, turn it all red in post, then selectively un-turn it red.
But if that’s all you’ve got and you can’t re-shoot it, you’re going to need to manually rotoscope the lenses. Resolve doesn’t have the strongest feature set for this. It *might* be possible, depending on the complexity of the shot (any motion blur, depth blur, hair, transparency, lens artifacts?), but other tools like Mocha pro would improve your odds.
So if it were me, and re-shooting wasn’t an option, I’d try to rotoscope in Mocha, import separate masks into resolve, and use those to change the color.
As far as layer nodes vs parallel, I don’t think it really matters as long as it works for whatever you’re trying to do. Test it first with simple shapes. Try the key mixer node to combine masks.
*and I really mean one narrow wavelength illuminating the entire scene. As soon as any other color enters the picture, you’re adding some kind of data to the green and blue photo receptors.
Sent from my iPad using Tapatalk
-MacBook Pro (14,3) i7 2.9 GHz 16 GB, Intel 630, AMD 560 x1
-[DR 17.0 Beta9]