Tom, this looks quite promising!
Any chance that you can set up a static subject and shoot for us a side-by-side set with no polarizers and varying cross angles, say 45, 60, 75, and 90 degrees? Maybe some sand, a dead fly, and a chunk of leaf?
I am particularly struck by the increase in saturation that you point out. It seems especially obvious on the eye. This makes sense to me (after considerable thinking!) because the eye is actually composed of a bunch of very small shiny facets. With normal lighting the eye is always giving us a substantial amount of light from reflections that are actually specular, even though they are individually so tiny and spread out that they look diffuse. Since specular reflections are non-colored, the saturation drops.
I just now ran some quick experiments using sheet polarizers with a flashlight in a dark room. The increase in color saturation depends on the material. This particular room was a storage area with numerous cardboard boxes. There were significant differences even from box to box. Some of the cardboard & writing showed no difference as I rotated the polarizers; some other cardboard and writing became obviously more saturated. On closer study, it seemed that the material showing no effect had no shiny surfaces even under magnification, while the material showing large effect had some sort of shiny coating on the fibers.
I tried finding some literature about this and came up pretty dry. It's commonly written that using a single polarizer will increase color saturation under some conditions, and it's also written that using crossed polarizers will eliminate specular reflections. But using Google, I couldn't find anything written about crossed polarizers and increased color saturation. I suppose this is common knowledge in some communities, but I sure hadn't thought about it.
Thanks for developing this technique!
--Rik
PS. New users beware: As discussed in an
earlier thread, crossed polarizers won't do what you probably want with most bare metal surfaces. They go black or close to it. The reason is that with a metallic surface all of the reflected light remains polarized, whereas with non-metallic surfaces the specular reflections remain polarized but bulk reflections become non-polarized. If you want a quick demo, try the flashlight test with a penny.
