Digital BW, The Print

I've been experimenting with different ways to profile the C6 inkset in
the 2200.  The QTR rip appears to be just a bit better than the Epson
driver, but they are so close I don't think even with a magnifier anyone
will notice the difference.

With the Epson driver the LK cross-over can cause problems with curves
unless the curves converge evenly as they approach 100%.  I've been
using my Grand Canyon shot -- http://paulroark.com/Grand_Cyn.html
<http://paulroark.com/Grand_Cyn.html>  -- as the test print  Even when
the 21-step ramp looks perfect, most of the Photoshop curves I tried put
artifacts in the sky.  The artifacts look like haze waves and are caused
by the slope of the ramp being un-even.  So, it's a very tough image to
print well.  I'm quite sure the LK cross-over is what causes the
problem, and having the color curves converge very evenly solved the
problem.

While partitioned curves made in Photoshop and embedded in an ICC with
Create ICC-RGB make among the best prints, a very simple grayscale ICC
is visually equal and essentially perfect.  That is, I just print a
21-step test file with the driver at 2880 and Color Controls gamma 2.2. 
I read that 21-step file and make an ICC with Create-ICC (no curves). 
Then print the test print with that grayscale ICC, and the image is
visually -- even with a magnifying hood -- as good as the best I've
seen.  So, easy works for me.

1440 will be visually OK for most, but upon very close inspection, I can
see the difference between it and 2880.

Since I'll actually be using the 2200 for my printing, I'll no doubt
have a number of profiles for it with time.

Next up is the C88+.  Epson has them in stock, so I bought one for my
daughter who is going away to college and will need a printer.  While
I'll set that up with a SuperJet (eBay) CIS and color for her, before
that I'll see if Carbon-6 works for that printer.

Paul

www.PaulRoark.com <http://www.PaulRoark.com>





[Non-text portions of this message have been removed]