falotico - there's tons of chatter about this subject over on the DT forum. It isn't really the sharpest process out there, by any means, but
prolonged transfer times just slow down the entire printing session and allow excess dye bleeding. Some people even ferrotyped them right
after the last transfer to quickly dry them. The idea is to let all the dye transfer off the matrix. This can differ a little from dye to dye; but once it occurs, you then move on to the next color of the transfer. How much dye the respective matrices absorb is determined by the acidity of the baths and numerous other tricks which make the process so flexible. To my knowledge, Ctein worked exclusively in the pan matrix process. I haven't visited him for awhile, but even the look of those prints is somewhat different than those made by the conventional blue-sensitive process, which is to be expected, because the starting point is itself so different (chromes vs negs), and the controls in between are likewise different. But even the dyes involved can be specially tailored for specific subject matter - the ultimate example of this
is how different dye sets would be chosen for particular Technicolor movies, then they would match the decor all filming, and whole set design to coordinate with it. ... sometimes I get totally distracted from the whole plot line of some classic old Technicolor movie just by
observing the skill of the filming. Some dye transfer printers like Ctein stuck strictly with official Kodak dyes. Others fiddled around quite a
bit. This potentially affected archival considerations, but that's a relatively contentious subject in its own right.