This is a 16-bit Photoshop environment. Because I am summing two relatively bright layers, I have overexposed and clipped the photograph.
Once the values go over 65536 (256 in 8-bits), they are lost, and dimming them later cannot recover them.
If I drastically overexpose in a 32-bit composite, as I have above, I still seem to clip above white.
But I can still recover the 'lost' information.
Photoshop's 32-bit environment is different from its 16- and 8-bit versions. In 32-bits there are not simply more divisions between black and white, there are values far outside those limits.
I'm not sure why Photoshop doesn't do all its math at 32-bits internally and then just cast to 16- or 8-bits at output. I'd use the 32 bit environment more often, but for some reason not everything works in 32 bits.
UPDATE 5/11/2009

Leave a comment