Re: Gum Color Gamut & Beyond CMYK?
first we have to agree on a few things, the CIE colorspace is theorical or
if you prefer it is based on studies to figure out what our vision could
actually see and not necessarly what we can actually see. In the real world
of materials, especially in the case pigment, it is very unlikely that you
will find one that comes anywhere near the edge of the diagram specially in
the green area. The phospors used to present us this CIE diagram can't even
come close to these "real" color of this diagram near the edges.
In the digital world we obviously use numbers to represent the colors of
this diagram and we have to remember numbers are an abstract representation
and it's easy to find combinations of numbers that have no physical
equivalent on our screens and or on printed materials. We have basically 3
levels of representation to work with or to juggle with, the numbers used to
represent the image colors, the displayable colors on screen and the
printable colors. We could also add to this our own physical limitation in
the case of our own individual visual system.
I assume here we are all artist ("want to be in my case") and for me at
least this mean putting a more or less important part of ourselves in our
work. This part of ourselves as nothing to do with technical matters it is
just as abstract as numbers are and as such it can become anything we decide
it to be. This also mean that your way of expression and the technical means
you choose to realise it are just as good as anyone else choices, there is
no right or wrong way to express onerself.
In conclusion to all this above, on the technical side the key word is
"compromise" which translate to what the specialist call gamut mapping
(which I'm not by the way). I for one prefer to learn the technics of
"compromising" with colors from a theorical perspective such that I don't
have to do all the required experimentation myself, I can learn from others,
there is a lot of work that has been put in this color world of ours. I find
this approach more efficient both in time and in cost. There is a big but in
all this, that's my own individual choices and my rational for choosing this
Now, how would I try to reproduce accuratly green 520 on or very near the
edge of the CIE diagram using pigments, the simple answer is I wouldn't even
try. At the level of my understanding it would be very difficult if not
impossible to find a material that would either reflect, transmit or even
emit a pure 520 nm frequency. From an artist point of view, assuming PS
tells me the color I want to use is out of gamut, I would either accept how
this color is map on the print using whichever technique I used or I would
choose some pigment that properly represent the color I'd like and use a
technique that would print it the way I want (hopefully), knowing in advance
it's a compromise and one I can live with.
----- Original Message -----
From: "Jacek Gonsalves" <firstname.lastname@example.org>
Sent: Sunday, September 21, 2008 11:06 PM
Subject: Re: Gum Color Gamut & Beyond CMYK?
> Thanks for the link David, that really is quite intresting. Will
> contact herbhoover and ask him how he did it...
> Yves, I understand that on a monitor screen there are always out of
> gamut information being displayed. Though its a matter of
> understanding between what color space your monitor holds and the
> image your using.
> My own initial creativity stems from the digital image and gum
> printing is the final product, that I want to retain the same detail
> and color(dependent on different color space of course) from my
> digital file.
> The above link shows a CIE colorspace, what the eye can see.
> Say I want to use a color at very top end, the green part where 520
> is. How could I do it? For a print I would need to find a pigment that
> matches that green? Come up with additional separations? Though you
> could argue that the image I use needs to have that green in it. This
> is where photoshop comes in, where you can tweak or increase the color
> to match that 520.
> I'm more intrested in matching my colors between what is on screen and
> what is the result on my print.
> Though in some ways I tend to agree with Katharine, that there might
> not be any reason to go a 6 separation route, if you were to choose
> vibrant colors in your 3 separation, CMY or RGB mix, I could possibly
> get that 520 green...
> A tool to find the exact the colors from screen to print is Datacolor
> Spyder printfix tool. For around $300-$400, you can get LAB and
> Density values from your print. Therefore I am able to plot what color
> space my prints fall under.
> ----- Forwarded message from email@example.com -----
> Date: Sun, 21 Sep 2008 07:16:23 +0100
> From: firstname.lastname@example.org
> Reply-To: email@example.com
> Subject: Re: Gum Color Gamut & Beyond CMYK?
> To: firstname.lastname@example.org
> Take a look at this photostream in flickr
> It's exaccerlly what you're talking about
> David H
> On Sep 21 2008, Katharine Thayer wrote:
> See my page on tricolor gum printing
> for a brief consideration of this topic (scroll down the page to the
> subheading, "Is it possible to reproduce the entire spectrum using
> three color layers?" )
> The answer, from Bruce McEvoy, is no, you do need to add the
> secondary colors (in other words, in gum printing, use six layers
> including the secondary colors in addition to the primary colors, in
> order to produce accurate colors throughout the spectrum, and I
> suppose some sort of profiles as Jacek suggests might be the way to
> generate the color separations for the secondary colors. But this
> seems to me an enormous effort and possibly a misplaced effort to
> boot; if you're looking for this kind of color precision, gum may not
> be the right process for you. Not that I would discourage anyone
> from dedicating themselves to this kind of study if it appeals to
> them, just that for me and my house, I don't quite see the point.
> However, the purpose of the extra colors and separations, as I
> understand it, is not to extend the gamut so much as to refine it.
> Depending on the pigments you choose as primaries, the layered color
> mixtures will produce different palettes of greens, oranges and
> purples; generally when printing tricolor we pick three primaries
> that give us the kind of secondary mixtures we want. For example, I
> generally prefer ultramarine or prussian blue for tricolor because I
> don't like the greens that pthalo produces with almost any yellow,
> they look unnatural to me, and since I print a lot of landscapes with
> trees/leaves, I want a more natural looking palette in the greens.
> Adding the secondary colors to produce six layers will theoretically
> make it less necessary to pick and choose primary pigments to produce
> the kind of secondaries you want, because you can choose the
> secondaries to your liking. But like I said, this seems a heck of a
> lot of work, when three colors, chosen carefully, will almost always
> give us a palette that gives us a reasonable enough approximation of
> realistic color, with special attention to producing the secondaries
> that we're most interested in.
> On Sep 20, 2008, at 9:11 PM, Jacek Gonsalves wrote:
> > Hi all,
> > Just continuing what Loris was talking about in regards with using a
> > predefined set of pigments and building a profile for Gum printing.
> > If one could build the profiles based on the CMYK pigments, can we
> > say that this holds a certain color gamut. Lets say we want to
> > increase the color gamut?
> > There are other methods as Pantone Hexachrome method of 6 color
> > separations CMYKOG, adds orange and green separation. Adding these
> > additional separation would therefore increase the color gamut. That
> > is if the image you have includes those additional colors.
> > I'm sure there are methods to make additional separations in PS,
> > though it might be complicated? If someone knows how to do this
> > please chime in..
> > Though it excites me to think that one could increase the color
> > gamut from finding what pigments fall further than the usual inkjet
> > printers do these days.
> > Regards
> > Jacek
>  http://www.flickr.com/photos/herbhoover/422834327/
> ----- End forwarded message -----