Dr Abhay Sharma is a Color Consultant from Abhay Consulting and he is also a professor from Toronto University. He started by saying that we need numbers to measure colour, and he doesn’t understand why the industry uses Yxy. Many industries use LAB which is a better colour space for understanding human colour perception.
Humans can’t describe colours easily and it is very difficult to remember colours to an absolute standard. Humans get tired and are inconsistent and this affects their colour perception. People are also colour blind. That’s why you need measurements and specifications.
In a side by side comparison, our human colour vision is probably better than equipment, but not if the colours are not seen side by side – many factors can affect how you see a colour.
In busy images, colour changes and inaccuracies get masked, but in images that have large areas of colour, the inaccuracies seem clearer. We’re good with neutrals and flesh tones. Sharma showed some good examples how the human visual system is fooled by colour casts and also the purity of light sources.
Instruments today incorporate a numbering calculation that takes into account of the human visual system, the light source and a sample that is being measured.
1931 CIE Yxy, which is usually used for characterising displays, is a very old standard. 1976 CIE u’v’ is a more modern version that is more perceptually linear. CIE L*a*b was also introduced in 1976 and is considered by many to be a better system.
He explained the non-uniformity of the Yxy diagram and explained that is why other industries have moved away from using it.
Colour is three dimensional with lightness, saturation and hue as the key parameters and this is reflected in L*a*b*, but not in the two dimensional 1931 Yxy system. Sharma showed how viewing the difference between a DLP and LCD projector in a L*a*b made the differences clearer. He called for the industry to adopt CIE L*a*b* as the way of specifying colour.
In questions, there was a lot of discussion of the issues of measurement and metamerism, especially in laser and laser-pumped phosphor.
Analyst Comment
Sharma didn’t really give a concrete reason for why L*a*b* should be used other than that it is “better” and more accurate. That is a worthy aim, but is unlikely to be a compelling factor in changing such well established behaviour (BR)
