Hello Fabrizio, I also did a lot of experiments with FOGRA39L and different separations/black generations. Basically I like the Argyll-generated separations more than the ones I get using the original profiles. Even if some primaries appear a bit more dirty than the originals, my separations are allways very smooth so that a color correction (even a default one) is not a problem at all. About the black generation / TAC: The examples you posted -were they generated before or after calculating a profile? I have the expirience that my curves become smoother after generating a profile (using parameters witch I test with xicclu -v -g -fif -kp......) and than calling xicclu -v -g -fb -ip <profileName.icc> Depending on the profile quality (medium or high), the curves may even become more smoothened. My "old" way of testing the separation quality of a profile is simply in generating a Lab-ramp (from 0 0 0 to 100 0 0) and converting using perceptual ri. This what you see in your separations responds almost exactly to what you will get running xicclu -v -g -fb -ip profile.icc Argyll also offers a few excellent images of testing profile conversions. Regards Vladimir 2010/4/1 Fabrizio Levati <fabrizio.levati@xxxxxxxxxxxxxx> > Hi Vladimir, > > > > Maybe it would be good if Fabricio could tell us for which printing > technique > > the profile will be prepared for. > > I was trying to create a profile starting from FOGRA39L with the purpose of > limiting the TAC to 240% or less. This profile is intended to be tested in > re-separating ads files before printing them on newspapers. > > > > If it is for offset/weboffset printing, Than > > I don't see the reason of limiting the TAC to 220% at the same time while > > limiting the black down to 92%. > > Even if you would manage to get smooth curves > > with that setting, I rather would question the quality of your final > print. > > 220% is OK but IMHO the black should not be under 95%. > > No objections at all! I started with TAC 240% and 95% black, saw that > playing with the total ink limit caused the curves to become rough at the > edge and get curious... > > Fabrizio >