digi Mike: matching single light references wip take 2

Howdy!

So, a few team members are speculating that the references are of inconsistent birghtness. I was wondering the same thing about the reference image at some point but decided to trust and move on. Now to double check, I reconstructed the three gray chips on the swim cap as closely as possible based on 6 camera angle and guesstimate 0.31 float diffuse so we have something to align to across cameras. Interestingly, it turns out if the front camera (camera 3) is the anchor with ISO 640, I have to tweak the ISO for other cameras, from 550 to 750, to match the gray chips under the same lighting. This is not what I expect as they should be shot with same camera settings.

So, I rendered an updated contact sheet with ISO adjusted so the gray chips matched more consistently. I can sense some consistency across cameras now, except camera 05 where the chips are much darker from some shadowing or vignetting(?). The issue is the renders are both warmer and cooler than the references depending on which camera angle we examine, which got me confused. I've attached a/b wipe video for each camera if you are interested in taking a closer look. Are these indicating:

1) spec falloff is off?

2) diffuse map color is off?

3) something else?

What is your thoughts?

cam 01 A/B wipe:

cam 02 A/B wipe:

cam 03 A/B wipe:

cam 04 A/B wipe:

cam 05 A/B wipe:

cam 06 A/B wipe: