2006.12.05 22:28 "Re: [Tiff] Grayscale, or is it?", by Edward Lam
when you read 8bit MinIsBlack greyscale (Photometric 1, otherwise known as BlackIsZero and PHOTOMETRIC_MINISBLACK) from TIFF, in order to display on a normal sRGB monitor, do you do any non-lineair conversion in between decoding and display, or do you just set the R, G, and B values directly with the greyscale values as read from the TIFF file?
When I read the spec as you've presented it, I now think option 2 is more correct. I'm not dogmatic about it either way.
In practice, I'm lazy and never do implicit gamma correction anyhow, relying on the user to decide. (I'm not trying to start another debate here so don't shoot me if you think I'm evil incarnate. :) Everyone should be using linear 32-bit floating point files these days right? :)