http://www.youtube.com/watch?v=sTxaHhMYSYE

Interesting video; I like that they emphasize the need to consider 2 gamuts: a proofer gamut (A2B1 Tag: device coordinates ->Lab direction) and a rendered gamut (B2A1 Tag: Lab -> device coordinates). The original poster of this thread mention 'gamut volume' as a quantitative measure of quality, but that criteria has to considered judiciously. For proofer gamuts, bigger (greater volume swept out by the transform) is always better; it should be a reflection of the native gamut of the hardcopy device that was characterized by measurements. But for the rendered gamut, bigger is not necessarily better - ideally, there is an optimal volume that lies between max volume and a much-too-small volume. The two gamuts should match up above the cusp (max chroma at given hue angle); but below the cusp, the rendered gamut should be somewhat inside the proofer gamut if the profile is any good. There are basically two reasons for this; the less important one is that constraints (TAC, ink limits, etc) imposes a smaller gamut. But the more important one is to preserve smoothness and continuity in the inks as you move toward the shadows. If you maintained a max gamut below the cusp, then small variations in color (Lab space) would result in large variation in inks. This would wreck havoc with shadow detail. Harold Boll ------------------------------------------ "Keep it simple, but no simpler!" ------------------------------------------