The source transparency and the destination old-style black and white print were both measured in Lab. refine was not quite doing it (deltaE average about 15) so I was to convert source Lab to RGB and use colprof that reached excellent results with peak err = 0.159433, avg err = 0.011282 on verification with a different part of target, not used for profiling. Next, I was to proceed using other tools to create the necessary device link. Result http://cl.ly/0t1d142Q443t/gfprint_abst.icm.zip -- Iliah Borg ib@xxxxxxxxxxx