[rmuhlack (Richard)]

looks like a great start, and I can't see any obvious blooming there either. is it simply dithering that has brought the blooming under control (together with a sigma reject / median stack) or are you using some other tricky method