Re: colorsync-users digest, Vol 2 #440 - 6 msgs
Re: colorsync-users digest, Vol 2 #440 - 6 msgs
- Subject: Re: colorsync-users digest, Vol 2 #440 - 6 msgs
- From: Jack Bingham <email@hidden>
- Date: Sat, 28 Jul 2001 06:59:42 -0400
- Organization: Jack Bingham Studio
Recently we have discovered a problem with some ColorChecker DCs when
used in polarized light situations. Specifically, some of the gray
patches in the I8, J8, K8, and L8 set exhibit a severe change in
lightness when exposed to polarized light. This problem does not occur
with non-polarized lighting.
So you can use one profile for different lightning
conditions (best results naturally when you profile each lightning)
to get usable results.
It is important to remember that profiling a camera is a process of determining the
camera's color capabilities not how it responds to a given light source. All we
care about is the camera's range. Once you gray balance under any lighting
condition you have linearized the camera to work with a single profile. Furthermore
since how the target is lit, perfectly even brightness and color temp, determines
the accuracy of the profile you can not just drop the target on the set and shoot
it. Imagine if one side of the target has a slight shadow cast across 10 patches.
The data from those 10 will be completely invalid. On the subject of the DC color
checker, I recently did an extensive test using the original color checker, the dc
and a Kodak IT8. The results were very clear. The original color checker makes a
much better profile hands down.
I would assume because the dc has different dyes and materials in the target that
they would naturally respond in a non-linear fashion to polarization. I would make
my profile without polarized light using an old color checker. Again it does not
matter what the lighting does because you are going to gray balance under those
conditions. All you need to know is the size of the color bucket that the camera
can fill!