RE: Gum Tri-Color Yellow

From: Keith Gerling ^lt;[email protected]>
Date: 07/09/04-02:53:59 AM Z
Message-id: <>

Bill says:

"The conversion to RGB from CMYK is simply a case of inverting the CMY
Channels to each of the RGB channels. There is nothing magic in it."

Actually that's not correct. Maybe in theory is should be, but in practice
it doesn't happen. CMYK (or CMY - and CMY is just CMYK with an empty "K"
channel) is not simply a reversal of RGB. Rather, it is a set of
instructions to the printer, which include percentages of ink to use, and
such arcane matters as "dot gain". If one considers Photoshop, for example,
the CMYK settings are practically infinite. Even the type of ink is
factored into the RGB to CMYK conversion.

"RGB only cleanly goes to CMY"

Not true for the same reasons as above. The RGB and LAB spaces are
lossless, but conversion to CMYK will always result in a changed file.
Well, almost always. I guess it would be possible to create an image that
is only one flat field of color, by scanning a Pantone chart, for example.

-----Original Message-----
From: []
Sent: Thursday, July 08, 2004 9:17 AM
Subject: RE: Gum Tri-Color Yellow

I've not tried RGB negatives, but I should think that they are functionally
equivalent to CMY negatives. You are using CMYK pigments, so in the places
where the Cyan should be in the print, the neg is clear and place where you
don't want Cyan the neg is opaque (and similarly for Magenta and Yellow and
Black). If you get there by taking a CMYK image, and inverting the channels
and making your negs from them, it or converting your CMYK image to RGB,
then taking the separate channels and printing them, I think they should be
effectively identical.

The conversion to RGB from CMYK is simply a case of inverting the CMY
Channels to each of the RGB channels. There is nothing magic in it. The K
channel is then applied to the other RGB channels.

The conversion back to CMYK, BTW, unless Photoshop simply maintains the
color information in CMYK and only uses RGB for some manipulations and
display, is not absolutely reversible, because you will loose information
going from CMYK to RGB to CMYK again. RGB only cleanly goes to CMY. The K
channel must be resynthsized out of the RGB information, and it probably
will not be identical to the original.

Assuming the values of the colors are in a range of 0 to 1 where 0 means no
color and 1 means full color, simplistic views of the arithmetic conversions

R = 1 - C - K
G = 1 - M - K
B = 1 - Y - K

we subtract the K value because K makes things blacker, and when R, G, and B
are 0, they make black. If as a result of these calculations R, G, or B go
negative, they are set to 0.

Converting back, we get

C = 1 - R
M = 1 - G
Y = 1 - B


This step is called black generation is is used to synthesize black out of
the CMY.

C' = C - K
M' = M - K
Y' = Y - K

This step is called undercolor removal to avoid saturating the paper with
excess ink and overprinting too much. This comes from printing presses to
prevent soaking the paper, but the calculations and process still hold in
modern computer printers.

If you take these calculations and work thru them with some set of values,
you will see that you don't necessarily end up with what you started with.

To answer Giovanni's question, if you wanted to be scientific about it, you
could probably use the eyedropper in PS, determine the color you are seeing,
and using these calculations to figure out how much more of any particular
color you might need.

Finally, the K in CMYK stands for blacK, not Key. This is used because B was
already taken for the Blue in RGB.

If you want more detail, the Adobe PostScript Language Reference manual, 3rd
edition, page 210, contains much more information than you probably EVER
wanted to know about RGB, CMYK, and several other color spaces.

Bill Leigh
> I stand corrected for sending such a muddled message :$
> What I really meant to say was if you print CMY pigments with RGB negs,
> in other words, C with the R channel neg, M with the G and Y with the B,
> is the result (in the opinion of others) better, worse or equal to
> printing the CMYK set of negs (all three colours plus black)? I do this
> regularly and it works for colour balance and appropriate tone, although
> I'm not happy with the overall contrast. I'm not sure if this is due to
> my curves not being right or my printing processes.
> Supposedly ( and this is what I've read) using the RGB set of negatives
> means that the contrast appropriate to the image is carried in the
> colour layers, and when you print with CMYK the contrast is mostly
> carried in the K (Key) layer. So you shouldn't print with CMY negs
> without the K. Or am I just completely wrong??????
> -----Original Message-----
> From: []
> Sent: Thursday, 8 July 2004 8:46 a.m.
> To:
> Subject: RE: Gum Tri-Color Yellow
> A minor terminology point: RGB is used for additive coloring and CMY[K}
> is used for subtractive. Monitors are RGB, but printed matter is CMY[K}.
> Printed things look black becasue the printing primaries of CMY
> "subtract" colors of the other frequencies from the reflected light by
> absorbing the other colors. When you see red printed matter, you are
> seeing what is left over after the printed matter has absorbed the other
> frequencies. If you are doing tri-color printing, you are using CMY, and
> CMYK is CMY with a optional K to give a true black over the combination
> of CMY.
> --
> Bill Leigh
> >
> > How many of you gum printers use four separations? I have read that
> the
> > K carries much of the contrast and therefore screws up colour balance.
> I
> > was just wondering who prefers RGB, who prints with CMYK??????
> >
> > Kate
> >
> > -----Original Message-----
> > From: Katharine Thayer []
> > Sent: Monday, 5 July 2004 1:25 a.m.
> > To:
> > Subject: Re: Gum Tri-Color Yellow
> >
> > Katharine Thayer wrote:
> > >
> > > wrote:
> > > >
> > > > I wonder if the muddy brown is an issue not so much of the colors
> > used, but perhaps poor color seperation in the negatives, or too fine
> a
> > screen in the seperate negatives. Larger dots of each color might give
> > better color rendition, just as it does with inkjet printing and matte
> > papers.
> > > >
> > > > Mark Nelson
> > >
> > > Not likely in my opinion. I haven't printed tricolor from continuous
> > > tone negatives, although Dave has (sorry, I can't think of Dave's
> last
> > > name at the moment, but you know, Cowboy Dave) quite successfully; I
> > > don't remember seeing any brown tones in his tricolor gums.  I've
> > > printed tricolors from many different types and resolutions of
> digital
> > > negatives, and I've never seen this brown tone in my own tricolor
> work
> > > from beginning to end. So I'm inclined to say it has little to do
> with
> > > the negative and much to do with the pigment. The fact that it
> didn't
> > > improve for Tom with different curves but it did improve when he
> > > switched from PV19 to PR209 should be an indication right there (I'm
> > > assuming he was using the same negatives).
> > >
> >
> > This is not to say that I don't believe it's possible for someone to
> > make really eccentric color separations that create weird color
> > combinations, but why would anyone do that? Even if their intent was
> to
> > create weird color combinations; it seems to me that weird color
> > combinations can be better (and more cheaply both in time and money
> for
> > negative materials) effected by choosing pigments that would create
> > weird colors in combination with eeach other than by trying different
> > weird color separations to see what happens, unless it's the
> > unpredictability that interested one.
> >
> > Also want to make it clear that I don't think this brownish or
> > brownish-purply cast has anything to do with how dilute or
> concentrated
> > the pigment is. I've printed with pigment concentrations ranging from
> > totally saturated color to just a whisper of color for each of the
> > primaries, and have never seen these brownish tones with the pigments
> I
> > use; the color combinations are clear and not brown no matter what the
> > pigment concentration.
> > Katharine
> >
> > ---
> > Incoming mail is certified Virus Free.
> > Checked by AVG anti-virus system (
> > Version: 6.0.714 / Virus Database: 470 - Release Date: 2/07/2004
> >
> >
> > ---
> > Outgoing mail is certified Virus Free.
> > Checked by AVG anti-virus system (
> > Version: 6.0.714 / Virus Database: 470 - Release Date: 2/07/2004
> >
> >
> ---
> Incoming mail is certified Virus Free.
> Checked by AVG anti-virus system (
> Version: 6.0.714 / Virus Database: 470 - Release Date: 2/07/2004
> ---
> Outgoing mail is certified Virus Free.
> Checked by AVG anti-virus system (
> Version: 6.0.714 / Virus Database: 470 - Release Date: 2/07/2004
Received on Thu Jul 8 14:51:35 2004

This archive was generated by hypermail 2.1.8 : 08/13/04-09:01:11 AM Z CST