I try to avoid profile editing at all costs. If a profile isn't perfect to begin with I'd rather get to the root of the problem than put a band-aid on the problem (ie: edit the profile).
[a href=\"index.php?act=findpost&pid=193961\"][{POST_SNAPBACK}][/a]
When I edit a profile, it has always been to improve the Soft Proofing tables, never the PCS to Device side of things. Profiles I've made myself (with Color Munki) and others I've had made (local Pictureline, Cathy's) have printed output that all comes out "reasonable" (e.g., no clearly wrong shades, no tone reversals in a grey scale step wedge, ...). It's always seemed to me that if I could just get the soft proofed image (on the monitor) to "acceptably" match the print, then I could proceed with confidence in making final corrections to the soft-proofed image I see on the screen (resulting in a print that matches what I saw closely enough that I don't feel compelled to change something and re-print).
However, I've only been at this for a year, and could be looking at things in the wrong way. Perhaps I am expecting too much from the soft proof. It would be great to see, on a single canvas, a test image, and a re-touched version that represents what an expert would judge as an "acceptable" difference between a soft-proofed monitor image and the printed output (perhaps the re-touched part is a bit less saturated, maybe the yellows are a bit greener, whatever).
I do try to standardize the lighting conditions for viewing a print, using a home-made, adjustable brightness 6500K light box (parts cost ~$400). But maybe, as I said, I'm expecting too much.