Digital BW, The Print

>> ... 1160 and I'm ready to go with a new inkset.  

>Unfortunately, the UT2 inkset will not work on the Epson 1160.

>Before using the UT inkset I used the FS-N inkset and would not go 
>back. The UT inkset has a better longevity rating and my printer is 
>now pretty near clog free.

>But, maybe Paul has another inkset to replace the present FS-N inkset 
>that would work on the Epson 1160?

Yes, there I've formulated "UT-FSN" and "UT-FS-Warm" (carbon) inksets (as
well as intermediate tones), but MIS needs to get them on the website before
they'll be available for sale.  (Don't bother calling MIS before then.  They
can't process the orders.)  

The UT-FS-W (carbon) is warmer than the current FS or the old VM warm.  For
a "neutral-warm" tone/hue mix 1/3 warm with 2/3 neutral.  For warmth equal
to the old VM warm curve, mix 1:1.  I'm not sure if MIS is going to have
these intermediate tones pre-mixed.

I have new curves for the 1160 (and 3000) for the UT-FS inks.  Even though
they are the same densities as the old FS (& Piezo) inks, I wanted to re-do
the curves.

I am not a fan of the old partitioning curves.  Some partitioning can't be
avoided with the 3000, and the 1160 will still give very slightly smoother
highlights with partitioning.  However, the radical partitioning curves were
the source of lots of the posterization problem.  They fight the Epson
driver's built-in partitioning, and they accentuate the production
differences between printers.

So, for the 1160 in addition to fully partitioned curves for EEM and
PhotoRag, I have a grayscale curve and a "mild" RGB partitioning curve
(RGB-EEM-2).  

This mild partitioning curve, I think, hits a good compromise between
smoothness and not interfering with the driver's attempts to control the
inks.  It should be relatively tolerant of printer differences and easy to
modify for different papers.  All the individual color curves start at the
(0,0) and end at (256,256), and there are no reverse slopes.  The approach
puts predominantly light inks in the highlights, then has those curves go to
a very low slope in the midtones, where the dark gray (cyan ink, red curve)
goes into a nice even and manageable slope. 

I hope that this approach solves some of the partitioning problems.

Another thing to try whenever you see vertical lines in the smooth ramp of
the 21-step test file (don't just look at the steps -- those mild looking
vertical lines are trouble) or you have some flat spots in a print is to
convert the RGB file to 16 bit before applying the curve.  Once the curve is
applied the file can be re-converted to 8 bit for printing.  Whenever the
main curve that is laying down ink has a low slope, the Photoshop curves
algorithm seems to need the 16 bit mode to avoid stair-steps.

Paul
www.PaulRoark.com 

For UT2 information, curves, and settings see:
http://home1.gte.net/res09aij/