As an AP, I never deal with skin tones. How difficult is it working with skin tones in post?
Well it depends on many factors, how bad, or far from the final output the original image is.
There is a lot that can be done in post, but in fact, if the original does not stand still, very little can be done.(or very expensive)
You can train some horses to do racing, but it will not work if you take a donkee for that purpose.
So post can't replace a good set.
And, oh yeah, the MUAs...and SPECIALY the MUAs with motion...because this has big consequences on the time to have to spend on
trying to recover bad make-up.
I find that the most difficult is not specially to lite well, but to lite with a final look in mind.
And that part is none of my business-habilities so I can't comment on that.
In general, in motion, the "rule" is to shoot the flatest possible files with log, to preserve the max range so
the looks are generated in post.
Now, someone like Coot who know how to lite, may not use log but directly lite to the closest final look in mind.
You can even create a look from RedCine X and load the metadatas into the Cam and shoot with it.
The Raw image is preserved so if you're not happy with it, you can always revert and recreate as many looks as you want.
That has an advantage if you know what you are doing, but a very big downside if you don't nail it because then the
footage is not really easy to recover. Using log flat file is adding extra steps in the pipeline but more flexible and safer.
The prob shooting log, is if you have clients that don't know motion, they look at the footage and go crazy because it
looks horrible. If you are shooting motion, within motion and for motion, then people are aware and know the workflows.
So shooting flat with clients that aren't coming from motion world is a mixed bag.