Re: Digital Negs - RGB vs CMYK

From: Katharine Thayer ^lt;>
Date: 12/04/04-04:24:43 AM Z
Message-id: <>

Joe Smigiel wrote:
> >>> 12/03/04 9:58 AM
> ..
> So I think the original question, posed by Joe S., was whether you could
> get
> true color gums in CMYK as opposed to RGB...
> Chris >>>
> Well, not quite. I made the statement that I believed more accurate
> color could be achieved using CMYK separations. I later commented that
> Livick's work was the most accurate I know of.
> Then there's the thang about whether "realistic" equates with
> "accurate". I certainly think you can get realistic/believable color
> with either RGB or CMYK separations, but accuracy is another matter.
> Most realistic-color gum images I've seen (primarily on the web) I would
> not call accurate. Many tend to have a cyan/blue bias IMO.

This is an interesting observation; I've been puzzling about it since
yesterday. If most realistic-color gum images have a blue cast (I hadn't
noticed this, but I don't have a particularly sharp eye for color
accuracy so I'll happily take your word for it) then you would think
that the answer would be to use RGB separations rather than default
CMYK, because the SWOP profile is weighted toward cyan in neutral areas.
I don't know if you had time to look at the web page-in-progress I
posted the other day, but it includes a neutral grey patch printed on
gum from RGB-CMY and CMYK separations. The RGB-CMY separations call for
printing equal amounts of each color, (the densities of the separations
are all 59%) and I got an almost, but not perfectly (slightly
pink-tinted) neutral grey from printing these separations in a
straightforward manner. You can see these results on the page, if you're

 The CMYK separations call for printing about the same amount of cyan as
the RGB separations (separation density 58%) but about 10% less magenta
and yellow (separation density 68% and 69%) The tricolor print I got
from these separations was not grey at all; it was very cyan/blue. Like
I said in another post, one could get a different result by printing
differently, but this was printed to hold everything but the separation
type constant, and this is the result I got, which is pretty much what I
would have expected from looking at the separations.
Received on Sat Dec 4 12:20:55 2004

This archive was generated by hypermail 2.1.8 : 01/03/05-09:29:43 AM Z CST