Digital BW, The Print

Frank,

You wrote:

>... I am ordering a CIS  for my  new 1160.
>Do you have any recommendation as to the
>Cone or MIS system being better
>for variable-tone work?

Short answer: I now favor the MIS version.

Long answer:

I tried to make them as interchangeable as possible. I want alternative
sources and competition.  The toners are essentially the same (1 part light
cyan to 1 part light magenta to one part lightest gray), and the ink
densities and positions have been matched on the 4-ink printers like the
1160.

So the differences in the variable-tone versions mostly relate to the
differences between the base inksets themselves.

MIS inks are slightly warmer than Piezo inks.  As such, the variable-tone
MIS version has a little wider useful range than the Piezo version.  The
control curves for the two version are slightly different.  The final
neutral-version prints, however, look very similar.  I'm not sure I can tell
them apart.

I took Piezo, MIS and variable-tone test strips with me on a trip through
the Southwest.  This trip included stops at U. of AZ Center for Creative
Photography and a number of high end photo galleries, where I could compare
the colors of these inksets to many classic prints and modern silver and
platinum fine art prints.  The MIS native color is very close to most of the
platinum prints I saw.  The Piezo native color was not seen in any classic
prints.  I saw one photogravure that was close.  However, usually the Piezo
print color (all test strips were on Archival Matte) looked too green.  This
color difference seems to affect mostly the warm versions of the
variable-tone prints.  In fact, the green is reduced even in the warm
variable-tone Piezo because the Piezo yellow ink is not used nearly as much.
The green seems to come from dyes in the magenta and yellow position inks.

Of course (not to avoid the obvious difference) the MIS inks are much less
expensive than the Piezo inks.

In my tests, the MIS inks and the Piezo inks faded and color-shifted at
different rates.

Let me briefly explain how I do these comparisons.  I have a florescent
light tester that uses a brightness way above what is normal in display.
The temperature is also somewhat elevated, but not hot.  How the results
relate to real-world display life is anyone's guess.  However, the results
do seem to correlate with the relative performances I and others have gotten
in "south window" testing.

I always made direct scanner (Epson 1600) comparisons of control test strips
to test strips that have been in the fade tester at the same time, together.
So, there may be lots of factors/variables, but I think I've equalized them
for at least the direct light-fastness comparisons between inks that have
been in the tester together.  I'm satisfied that the results are useful
enough that I base my decisions on it as opposed to the less controlled
"south window" tests.  (And life is too short to wait for Wilhelm.)  When I
graph my results, the patterns I see look a whole lot like the graphs RIT
gets -- see the MIS tests, for example.

The results of these fading and color shifting tests, comparing standard MIS
to standard Piezo prints on Epson Archival Matte, are as follows:

At 100 hours:  Piezo looked better because it had color-shifted (warmed) 26%
less than MIS.  However, MIS midtones had faded 26% less than Piezo.

At 200 hours:  The color shifting at this point favored Piezo by only about
10%.  So, visually, Piezo's edge was essentially gone.  Piezo midtones had
faded about 40% more than MIS midtones.

At 300 hours:  MIS at this point pulls ahead in all respects.  The Piezo
midtones had faded about 65% more than MIS, and the Piezo test strip had
also color shifted about 25% more than the MIS test strip. I didn't go any
further with this test.  I think the pattern is clear.

Visually, the color-shifting/warming is more of a problem than the fading.
Many will not notice the fading in actual prints.  However, I think they are
both a result of the same process.  When graphed out, the relative rates of
change -- fading & warming -- for a particular inkset look virtually
identical.

I think that this color-shifting/warming is a real problem.  However, I
think I can control (virtually stop) the color-shifting in the initial
warm-up period, whereas I don't think I can control the ultimate fading of
the inks.  Thus, MIS's ultimately better fading (and color-shifting)
characteristics are more important to me.

Where am I going with this next?  I have one test strip with MIS inks with a
special "counter-shift" mix added to the inks that is as cool in the
midtones after 400 hours in the tester as it was the day it was printed.
I'll focus on finalizing this mix soon, and it will be compatible with the
MIS version of the variable-tone mix.  No promises here, but I think I can
control the warm-shifting to a great extent.

(I'm leaning toward putting the mix in the toner so that the neutral and
cool prints will stay that way, but the warm version will still warm-shift,
thus allowing people to sun-warm their prints to achieve the warmer, almost
sepia-like, tone that some people like.  With this approach, those who want
the counter-shift advantages can just switch to the new toner, hopefully
without even changing to new control curves.  It should be easy even with a
CIS.)

So, bottom line, I started with Piezo, but now my focus is on the MIS
inkset.  Price, suitability for hextone printers, interest in MIS in mixing
the inks, and ultimate stability are major factors in this.  (I think MIS's
publishing it's RIT test results is also a real source of confidence.)

(Frank, good to see you on this forum - hope all is well.)

Paul
http://www.PaulRoark.com