fun with tricolor gum

Hi All,

It's a dark and stormy night, raining relentlessly and dismal beyond
belief, and things have been quiet for a while (though it sounds
like a lot of people have been writing to the list and not getting
their posts through). So I feel like putting my feet up on the
heater and waxing philosophic about tricolor gum.

The main reason I stopped printing tricolor gum about five years ago
was that it got boring to me; I felt like I was working on an
assembly line or something. I had it down pat, it offered no
challenge, the results were predictable. I liked the colors I was
using; I liked the results I got, but it was just .... boring, to do
it. I prefer exploring new challenges to doing the same thing over
and over, that's just how I am.

So five years have gone by and my work has gone other directions,
from very pale ephemeral "monochromes"(actually made of pale layers
of several colors) to gum prints on glass. In the last year I've
started printing some tricolor again, although quite different from
the tricolors I've done before. The buyer of one of these recent
prints said it reminded him of Turner; the buyer of another said it
made her think of Maxfield Parrish. (I always feel like I'm on the
right track when my work reminds people of painters rather than
photographers).

But regardless of how tricolor gum has figured into my own work, my
interest in the issues around tricolor gum has continued, of
course. I've had in mind for several years to print the same image
in tricolor gum, once in transparent pigments and once in opaque
pigments, to show the difference. I wanted a colorful image to use,
which is something of a problem because my color photographs are
usually intentionally not very colorful. And then I thought of the
banana lady, Adobe's color test image, and one day earlier this week
I printed this image in transparent and opaque pigments (mostly
pigments fairly unfamiliar to me except for doing some step prints
with them last summer) using separations printed from my new Epson
1280 on two different kinds of transparencies: a 3M product that's
apparently no longer available, that didn't work on my old printer
but works great on the new one, and Pictorico OHP.

What with the unfamiliar negatives, the relatively unfamiliar
pigments, and the fact that the gum- pigment mixes I settled on with
my single-pigment step tests last summer turned out almost
invariably (except for the yellows) to be too heavily pigmented for
tricolor, (more about that later) all of a sudden I've had to learn
how to print tricolor all over again, and I'm having fun with
tricolor for the first time in ages. I'll show you my first trial
efforts, just for laughs and to show that even a seasoned gum printer
has to do some trial and error when you change everything at once.

These are just test prints; I wasn't trying to do anything serious
here. I was just dinking around to get the feel of it all; I wasn't
even careful to coat the entire image area. None of the test prints I
did is well color balanced, so I need to adjust pigment
concentrations and print at least one more print from each set of
pigments and adjust the pigment concentrations again if necessary til
I feel like I've got a better pigment set, before I print the final
comparison for my website. If I had time to do that now, I'd just
show you the final prints instead of these seat of the pants
explorations, but it will be a while before I can get back to this,
and I wanted to share some things from it while I'm thinking about it.

I also think I will try to choose colors that are more similar for
the final comparison; PR 209, for example. rather than PV19, to be
closer to the cadmium red. This is definitely a preliminary
demonstration. But, taking into consideration those caveats, the
prints made with opaque pigments are to my eyes duller; they have a
sort of film over them that dulls the colors, as a result of the
opacity of the pigments, whereas the transparent one is clearer and
brighter. This cloudy dullness may not be visible in the electronic
reproductions. As I look closely at the actual prints, say in the
face area, the opaque image appears to be made of intermingled dots
of the different colors, as if printed through halftones, whereas the
transparent one is a sheer blend of continuous color. I'm sure I
don't see how that would work, given that they were printed from the
exact same separations, but I can only report what I see. But the
opaque is very nice-looking, and as I keep saying, there's no reason
not to choose opaque pigments for tricolor. All I'm advocating for
is to be informed in your pigment choices.

Comparing the 3M separations to the Pictorico separations, the
separations themselves look very similar. The Pictorico negatives
don't have more ink on them; both types of separations are actually
quite thin, thinner than the separations I've used before, which were
either bitmaps printed from laser printers, or inkjet paper
negatives. You can see that, at the same exposure times, the
Pictorico holds back significant radiation compared to the 3M
separations printing the same coating mixes; the resultant print
looks quite washed out compared to the print from the 3M
transparencies, although the flesh tones are very delicate and nice,
and this is the only print where the pear is remotely close to the
right color. This print is also unfortunately fuzzy from imperfect
registration.

As for the issue of pigment concentration, I said here recently that
you can't just look at how the individual pigments print by
themselves, you have to look at how they print in combination with
each other, and in general you wouldn't want to print your cyan or
magenta as pigmented as a layer in tricolor as you would if you were
printing them as single colors. But I fell into the same trap, when I
was testing pigments for tricolor last summer. I started out printing
step prints of each individual pigment, then I was starting to print
combinations of two pigments, to show how each pair of pigments
prints the secondary colors (oranges, purples, greens) and eventually
I was going to print the third color on all of them to see if they
would produce neutral greys, but now, printing those pigment mixes in
tricolor, I find that in general, the yellows are okay as they are
mixed but the blues and reds are too pigmented for tricolor, and
these test prints are either too dark or have a color cast in neutral
areas or both, as a result. But I also have this other problem,
where I'm using this Daniel Smith gum that evaporates, or dries up,
or something, and so the mixes containing this gum become more and
more pigmented as they sit, and finally turn into a mass of pigment
that has a tarry quality and can't be resuscitated with the addition
of more gum; it has to be thrown out. So there are two things going
on there that can't be immediately separated out; at any rate, most
of my mixes that I had assumed were right for tricolor turned out to
be too pigmented, for one reason or another.

One of the pigment mixes was so overly pigmented that it gave a
partial "tonal inversion" when printed the first time; I added about
1/3 more gum to the mix before using the mix in a tricolor print, but
it's still too much, as you can see by the reddish cast on the print.

And then there's the question of color balance and color accuracy,
which is a different issue from the one I was considering, but which
becomes an interesting issue as soon as you start working with a
color test image like this; sooner or later you start wanting to see
how close you can get to the actual colors, just as a challenge. I
don't know how many of us will ever be confronted with an extreme
image like this, as far as the range of intensity of the different
colors. It's tempting to compare how the different pigments render
the different colors, but I don't think this is a useful exercise
unless there's a neutral color balance in the test prints, and in
mine there's not. So the fact that the cadmium red seems to render
the bright reds in the hat and the cherries more accurately than the
PV19, which renders them sort of maroon, may be related to the fact
that the colors aren't properly balanced, rather than to the fact
that the PV 19 is inherently bluer than the cadmium red.

   Whether accurate color balance is a goal for any particular
printer is a personal preference. I have never been particularly
interested in accurate color renditiion, in fact much of the appeal
of gum for me was that I could print softer, less saturated colors
than Kodak, and I went to some trouble to figure out how to print
that way. So I hope to goodness that I'm not going to get curious
and try to match the colors in the test image. It's very tempting,
as an intellectual exercise, to try to see if with one set of three
pigments you can match all the colors, but I hope I don't get
interested in that. But if someone else takes it on, I'll be
interested in the answer. And then there's the whole question of how
you can affect all that with curves, but I prefer to deal with it by
varying pigments and pigment concentrations, and in fact it doesn't
make sense to me to try to fix a pigment concentration problem with a
curve. If there's too much pigment, why not just add gum, instead of
cutting back on the exposure and wasting pigment? And if there's too
little pigment, or the pigment is too weak or light to hold its own,
a curve isn't going to help it. At any rate, It would be interesting
to see if you can achieve the same goal both ways, but as I say, I
hope I won't get interested in that myself.

The crucial thing for getting tricolor right is balancing the yellow
and the cyan. The yellow needs to be fairly strong, (here I used PY
184, cadmium yellow light, and PY 97; none of them really hold their
own with the strong concentrations of the other colors I used here,
although the PY97 comes closest) and the cyan needs to be held way
back, in order for them to balance properly, especially if you're
using pthalo, which tends to overwhelm the other colors if you're not
careful. The red (magenta, if you must) comes in between, but in most
cases it shouldn't be printed too concentrated either, as it is in my
print B.

Here's the URL; this won't stay up long because it's just for fun and
because I haven't asked Adobe for permission to use the test image,
but I will do that before I use it on a permanent page. It's possible
that I'm just talking to myself here; is anyone but me actually
interested in tricolor gum? I have the sense that most people are
more interested in the gumover kinds of combination processes. At any
rate, here it is, for whatever it's worth. Have fun,

http://www.pacifier.com/~kthayer/html/tricolorfun.html

Katharine
Received on Thu Jan 5 21:12:26 2006