AVS Forum banner
1 - 7 of 7 Posts

·
Registered
Joined
·
1,176 Posts
The more wide gamut a display is, the narrower the primaries, and hence the more observer variation is likely to be noticable. Even if personalized observer curve characterization was widely available, there's no way to make two displays with different types of primary spectral characteristics match exactly for all observers.


Bottom line - this is one of the fundamental disadvantages of colorimetric reproduction.


Any sort of multi-primary or spectral reproduction is likely to be expensive and is currently more in the nature of research rather than products. Such technology may never become mainstream, simply because of diminishing returns.
 

·
Registered
Joined
·
1,033 Posts
thanks for sharing this randal. I haven't yet watched the videos, but the white paper was very well written:


one quibble:


the paper reads:
Quote:
Note that this is not (or at least not primarily) a problem of metamerism, the well-documented problem of differences in the color perception among a population.

should read "observer metamerism", if I'm not mistaken.


Incidentally, I've always found that term a bit baffling. Observer metamerism, as I understand it, refers to a situation where the degree of metamerism between two light sources changes depending on the observers. If we all had exactly the same visual systems, without any individual variance, there would be no observer metamerism.


But, given the definition of metamerism, one would expect what is normally called "observer metamerism" to be called something like "observer metameric failure".


I also wonder whether the weaknesses of the CIE 1931 standard observer functions (e.g. the underestimation of our sensitivity to wavelengths under 460 nm) also are related to observer metamerism.


That is, are the areas of the standard observer functions that have been poorly characterized the same areas where we exhibit the most variation in sensitivity across individuals? This would make sense in a way, since it would explain why the 1931 functions are poorly characterized, namely they had a limited cohort and didn't measure across a wide enough variety of phenotypes.
 

·
Registered
Joined
·
1,176 Posts

Quote:
Originally Posted by spacediver  /t/1523445/what-every-calibrator-should-know-about-multiple-display-calibration#post_24510424


That is, are the areas of the standard observer functions that have been poorly characterized the same areas where we exhibit the most variation in sensitivity across individuals? This would make sense in a way, since it would explain why the 1931 functions are poorly characterized, namely they had a limited cohort and didn't measure across a wide enough variety of phenotypes.
Recent research has postulated that observers tend to clump into groups.

[ see "Identification and Assignment of Colorimetric Observer Categories and Their Applications in Color and Vision Sciences" and "Toward Reducing Observer Metamerism in Industrial Applications: Colorimetric Observer Categories and Observer Classification" by Abhijit Sarkar ]


Rather than attempting to re-formulate CIE 1931, the current interest seems to be more in the direction of creating parametrized color matching functions [ CIE2006 CMFs ], personalized observer CMFs, or at least being able to measure and characterize the observer clusters.
 

·
Registered
Joined
·
1,033 Posts

Quote:
Originally Posted by randal_r  /t/1523445/what-every-calibrator-should-know-about-multiple-display-calibration#post_24514354


The subject is more related to Spectral Power Distribution and how they differ within various technological architecture of displays. Did you observe the videos?

Haven't yet watched videos, but what we were talking about directly relates to SPDs. Some SPDs, particularly spiky ones that have peaks in areas of the observer functions that are poorly characterized, will magnify these problems.
 
1 - 7 of 7 Posts
Top