3

As I understand, on ordinary monitor I could only see a subset (e.g. sRGB) of the horseshoe shape of visibile color.

Is there any lab device that could generate all the visible colors and project them onto the CIE 1931 chromaticity diagram?

enter image description here

  • 1
    It's theoretically doable, but only as an emissive device and it would be incredibly expensive, It isn't even theoretically possible for reflective (passive) surfaces due to Macadam limits. – doug Feb 10 at 8:17
  • @doug how would you do it, even theoretically? All the points around the edge of the horseshoe would need to be a pure single frequency of light. – Mark Ransom Feb 12 at 5:15
  • @MarkRansom Of course. But one can approximate it well within arbitrary dE limits below perceptual level with much fewer single freq. sources. Since the xy chromaticity of two wavelengths follows along a line between the two on the horseshoe one just needs to increase the density where the curve is greatest. Probably doable with well under 100 specific wavelengths such that max dE would be under 1. It's obviously not practical but would certainly be possible. – doug Feb 12 at 5:48
3

In order to generate all of the chromaticity coordinates on the chromaticity diagram one just needs two sources, pass each of those sources through a monochromator, and mix the resulting two monochromatic sources in an integrating sphere. Any point on the chromaticity diagram can be reproduced by using the two wavelengths which are along lines that pass through the chromaticity coordinate in question and points on the spectrum locus where the line is perpetual to the spectrum locus.

One issue would be the luminance of the light at each chromaticity. Without a pair of equal energy sources the luminance of each chromaticity coordinate would be different.

It’s worth noting colored chromaticity diagrams are mostly color for sake of the aesthetics of the diagram only. Any chromaticity coordinate could appear as any color depending on the viewers state of adaptation. For instance, if a viewer were adapted to 550nm (unlikely but theoretically possible), “white” would be on the spectrum locus at 550nm and the vast majority for the diagram would be filled with shades of blue and red.

Regardless, one could easily produce any x,y chromaticity coordinate with just the setup described above.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.