On Thu, 8 Jul 2004 22:49:36 +0200, Jostein wrote: > So far, my impression is that the USM mask is less effective on > slightly soft images.
That doesn't match my experience, but I'm coming at it from film. Scanned at 4000 dpi, a lot of my film shows up what appears to be grain aliasing, which is sort of analogous visually to what you'd get from a CCD without the interpolation. That is, individual (or tiny groups of) pixels with strong color divergent from the color of their neighbors. I've found that running a light Gaussian Blur on the image before USM allows the USM to work a lot more effectively than without the blur step. TTYL, DougF KG4LMZ
