[argyllcms] Re: Of ink limiting and maximizing gamut
- From: Martin Weberg <martin.weberg@xxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Thu, 29 Jan 2009 01:31:04 +0100
On Thu, Jan 29, 2009 at 00:04, Alastair M. Robinson wrote:
> Here's a raw plot of a CMYK stepwedge testchart - plotting input against L*,
> a and b. Note that there's already a pretty hefty (modified) gamma
> correction going on just to bring the stepwedge testpoints to something
> approaching even spacing - maybe this is hiding effects that might otherwise
> be visible?
>
> http://www.blackfiveimaging.co.uk/linearize/EpsonPQI_raw.jpg
I prefer plotting a and b together looking through the L plane, which
I think better illustrates how CMY would effect the final gamut. If
the color bends (shift hue) really fast it's often really obvious that
adding further ink will not expand the gamut. Then I prefer to limit
right at the start of that bend or just a little beyond (subjective? I
know :-)).
> I have to admit I've been looking more closely at L* than a/b for signs of
> over-inking. How much of the a/b curvature here, I wonder, is down to
> over-inking, or does the fact that a given hue doesn't necessarily have
> totally constant hue angle in L*ab space have anything to do with it? (As
> per the infamous blue-turns-purple problem)
I don't know.
If I where to print step wedges or gradations like in many
illustrations (vector art) of primary and maybe secondary colors, I
would not be too happy if there was a visible hue shift. Then I would
sacrifice some of the gamut and inklimit before the visible hue shift.
When proofing I see no problem with a slight hue shift in primary
colors after inklimiting since all primary colors if needed will be
used to match the reference (rel. or abs. col.).
> I suppose the crux of my question is really why
> super-saturated-colorants-and-super-dense-black-but-low-ink-limit performs
> so much worse in terms of overall gamut than
> restricted-saturation-of-colorants-but-higher-ink-limits. I guess just
> that's an unusual scenario, and thus the tools aren't really geared up for
> it?
Maybe. I haven't run into this scenario before.
Martin Weberg
Other related posts:
