[argyllcms] Re: Neutralize Grey Axis

  • From: Ernst Dinkla <info@xxxxxxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Wed, 12 Jan 2011 15:05:05 +0100

Op 12-1-2011 12:49, Graeme Gill schreef:
Nikolay Pokhilchenko wrote:
Graeme, am I right in supposing that pointing viewers adaptation to
D50 for colprof
lead to better matchig of the grayscale prints with other true
neutral-in-D50 prints in
D50-illuminated enviroment?

Only if it's a more accurate description of the viewing situation. If the
print is actually neutral with respect to it's paper color, yet appears
to the observer, then the observer must be making that judgement with
to some other white reference. If they are viewing another print with a
warmer white point as a reference, then this would explain the effect.

A logical way of creating a better match is to specify what the situation
actually is (by overriding the default assumptions of viewer white point),
and then let the color appearance space matching do its stuff.

Graeme Gill.

What may work perceptually for a "full" color image could be less apparent for a B&W image. The highest reflectance spot or the most dominant light area within or near a multicolor image is likely taken as the neutral/white reference. In a monochrome image it could be the global impression or an average of the lighter tones.

There has been a somewhat heated discussion on the use of the term "metamerism" for what actually should be called color (in)constancy with changing lighting etc. Colorsync list half a year ago. As a side note I mentioned then that single B&W images, monochrome images, have some conditions that come closer to conditions set for describing a metameric match or failure: two samples as the starting point. We are much more aware of a b shifts over the tone range in images that ideally should have no a b shift at all along the tone range. It is much more likely that we use the most dominant a b value along the tone range as a reference to describe the deviation of the other parts and not use the media base as a neutral reference. So it is possible that someone describes the paper correctly as blue or the Dmax black as warm in a B&W image without another print or white reference next to the one described.

What I mentioned in the other message that profiling / media presets (LUTs) can not compensate a warm black ink in some cases is related to what I describe above. Papers with enough FBA content like the example b -11 mentioned in the thread will one way or another show too much hue deviation along the B&W tone range if the black ink is too warm. In the end you will at best get a kind of gradual split toning over the tonal range with lighting including UV. It will also be very sensitive to changing illumination.

Met vriendelijke groeten,      Ernst

New: Spectral plots of +210 inkjet papers:

|          Dinkla Grafische Techniek         |
|             www.pigment-print.com           |
|             (         unvollendet         )            |

Other related posts: