Re: UCR vs GCR

Dan Margulis <76270.1033@...>

Ruud writes,

So UCR and GCR both mean the same and are active in the neutrals (the L
axis in Lab).>>

It's clear that there is a lot of misunderstanding of this important topic
on the list so let me try to clarify what the terms mean and why they are

In RGB, all colors are unique. If you know the values that make a certain
color, there is no other set of RGB values that make the same color. In
CMYK, this is not necessarily true. 120r120g120b is a gray that can only be
produced one way in RGB. The equivalent in CMYK is in the vicinity of 50k,
or of 60c50m50y, or of any mixture of the two: the more black being used,
the less CMY. Therefore, one could have zero black, or 50% black, or 7% or
12.5% or whatever one wants, and still produce the same gray.

The same would be true of almost any color: remove CMY, and you can add
black. Even something as colorful as a face *could* have some black in it,
although it isn't customary. Only a truly saturated color, such as
70c50m0y, could not have black, because in order to add black, you have to
subtract CMY, and there isn't any yellow to subtract.

The basic tradeoffs in a heavier black is that you accept the risk of muddy
reproduction if either black comes down unexpectedly heavily or you don't
have a good handle on what black dot gain is. Also, it's more difficult to
color-correct images in Photoshop when they have a heavier black. As
against that, there will be fewer unwelcome changes in color, and if the
image is going to press, it will be easier to hold in register.

There is a tiny minority of fanatics who hold that one should always print
with as much black as possible (in Photoshop terms, Maximum GCR with
significant UCA). The remainder of the world realizes that all the benefits
of a heavier black are there if one only uses about half that amount in
colored areas. Going further, the majority of experienced CMYK
practitioners, but not all, prefer much less black than even that. This is
where the terminology confusion sets in.

Assume that you are a person who hates the idea of black ink and would just
as soon print entirely in CMY. That sounds like a reasonable philosophy,
but you have to make certain exceptions. Your deepest shadow will be
something like 100c95m95y. That will be too light and too red. So, if you
are sane, you have to add around 50k.

100c95m95y50k, however, is a total inking of 340%, which most printers
won't accept. Therefore, even though you hate black, you are forced to add
more of it here, so that you can lower the other three values and thus get
a lower total ink. Thus, addition of black, removal of the undercolors
(UCR=undercolor removal).

This is why in the type of image discussed by Hector, all the shadow detail
migrates to the black--the CMY channels have to be suppressed because of
the total ink limit.

For most types of printing black only necessarily will appear in dark
neutral colors, as Ruud suggests. However, for poorer types of printing,
such as newspapers, the lower total ink limit will force the use of black
in colors such as navy blue, even if the type of separation is UCR.

There wouldn't be much disagreement as to where the black would absolutely
*have* to appear. The problem is, confining black to only those areas isn't
workable. You wouldn't want to wait until 95c85m85y0k was reached and
*then* start adding black. The gradation in all four inks would be enormous
as the shadows got darker and the job wouldn't be printable. Instead, the
black has to start in areas that are at least slightly lighter so that it
doesn't get so dark so fast. And, of course, there is zero agreement as to
where it should start.

GCR--gray component replacement--means the use of even more black. But
again, nobody agrees as to how much more, where to start it, or how fast to
add it. If the software expresses GCR in a percentage, about all you know
is that 40% will give more black than 35% would. In Photoshop, "Light GCR"
means more black than UCR, and "Medium" means more than light. In practice,
"UCR" and "Light GCR" mean almost the same thing, "Medium" is much stronger
than "Light", and "Heavy" is a bit stronger than "Medium".

The advantages and disadvantages of a heavier black will be most pronounced
in subtle colors. For example, I just created in RGB a typical green for a
leaf, and separated it five different ways. The results are:

UCR: 60c20m81y1k
Light GCR: 59c18m80y2k
Medium GCR: 56c14m77y7k
Heavy GCR: 55c11m76y10k
Maximum GCR: 46c0m69y24k

No matter how heavy the black runs on press, it won't muddy up this color
if you are using UCR or Light GCR. Anything higher is a risk. OTOH, if your
agenda for the image is very subdued colors, having more black will help
insure that nothing goes wrong. In theory, all of these black generations
will give the same result. But as we do not live in a perfect world, in
practice they do not.

Dan Margulis

Join to automatically receive all group messages.