Feb
13
I’m playing around with matching colors using color charts and in theory it should work very well. And I guess it does, but I’m not really sure what the right process is. In the attached script, I a fg plate shot in a studio and a bg plate shot at a location with different cameras. I thought that by using the grade node and sample the black and white points of my original image color chart and then sample the lift and gain from the bg plate, would be the correct workflow. But it seems to give me better result by sampling the skin-tones of my plates instead of white. Why is this? How do you guys normally handle situations like this?
Again, this is just a quick test of concept, and it should be very quick to match colors with this method. I have seen the video on The Foundrys webpage about this, but he samples skin color there aswell.
Could you do something with the other colorson the chart and create a 3D lut somehow that would be very accurate?
I’ve attached the Nuke test script and the fg and bg plate.
Post a Comment