Digital BW, The Print

--- In DigitalBlackandWhiteThePrint@y..., "antonisphoto" 
<antonisphoto@y...> wrote:
> Martin,
> 
> there are different kinds of "profiling" in this regard.
> 
> The simple one, that deals with matching a grayscale file to a 
grayscale print 
> can be done right now in PS 6. Simply calculating dot gain (OK, 
that may 
> require an instrument, though eyeballing is mighty close) and 
saving the 
> resulting gray setting becomes a profile in the "eyes" of 
Photoshop. 

Antonis,

This is true but once you see a true color profile of a gray-
ink/paper combination where not only the density is reflected but the 
color of the ink and paper as well, it is really amazing.

> 
> Then there is the more complex one of dealing with separating the 8 
or 16bit 
> grays into the 4 or 6 inks. I believe you are already ahead of me 
on that one,

Just riding Tyler Boley's coat tails here!
 
> but my previous experience point to the necessity for a RIP capable 
of 4 or 6 
> color printing before any profiles can be made meaningfully. In 
other words 
> gray-to-CcMmYK would be preferrable to gray-to-RGB-to <unknown 
Epson 
> separation algorithm>.

Exactly, the Epson driver controlled C-to-c and M-to-m transitions 
seem to be very ink dependent. Not just a dilution issue but probably 
viscosity and surface tension related. In other words, Epson created 
their driver(s) to work with very specific inks and once you move 
away from those, you are on difficult ground. From Paul's success 
with the MIS VM curves and MIS's formulation of their inks it may not 
be of great importance but I am coming to believe that to extract 
that last 0.1% you may want to control these transitions.

If you have an accurate on screen profile, this seems to greatly 
speed and simplify the building of the separation curves.

> 
> But either of the above solutions still require a way to bring the 
hardware into 
> a repeatable, linearized, calibrated state. That means that for a 
given digital 
> value it has to produce the same exact density of ink on paper. A 
tall order for 
> most desktop/toy printers. It's probably why the wide-format Epsons 
cost that 
> much more. But even then, plotting graphs to check for drifting 
(what my 
> original message was about) should be a regular part of any 
serious  printing 
> "regiment".
> 

Exactly. Ideally you need something in your workflow that allows you 
to respond to the changes you measure and I don't really see this in 
Piezo or any of the current workflows. It is of course there in the 
RGB, CMYK and CcMmYK separation methods, but only in the cumbersome 
method of generating a new on screen profile and new separation 
curves. Separation curves seems as much art as science to me at this 
point.

Martin