CIE Chromaticity diagram: rogues gallery

by W.A. Steer PhD
Back to contents About...

You see many examples of badly-plotted (severely chromatically inaccurate) CIE chromaticity diagrams. This page shows what goes wrong, and why!

CIE Chromaticity diagram: rogues gallery

In the critique below I have left the out-of-gamut areas grey, since there is no technically-accurate way to show these colours. Any colour in this region can never be more than an aesthetic fudge.

This is perhaps the worst you'll see! In order to get the most brightness across the plot, per-pixel RGB have been scaled up so that the greatest of the three always equals 255, leading to the ghastly and misleading bright-Y artifact. Furthermore, the gamma-mapping of the display has been ignored - the RGB values naively being treated as if linear in light. As left, but at least gamma has been treated properly.

This doesn't use the RGBmax trick (the blue and green vertices are less than full brightness), but still has ignored the effect of display gamma. This image is technically correct, has the correct gamma, and is also free of the nasty bright-Y artifact. It is a bit dull however, as no alternative enhancement has been made to get the maximum brightness from the green and blue corners of the triangle.

And finally: this is my version, with correct sRGB gamma, and a linear luminance enhancement to get the green and blue corners to full-brightness for greatest appeal:

Return to Introduction to Colour Science page

Created: October 2004
Last modified: 26 October 2004
Source: http://www.techmind.org/colour/rogues.html


This is perhaps the worst you'll see! In order to get the most brightness across the plot, per-pixel RGB have been scaled up so that the greatest of the three always equals 255, leading to the ghastly and misleading bright-Y artifact. Furthermore, the gamma-mapping of the display has been ignored - the RGB values naively being treated as if linear in light.	As left, but at least gamma has been treated properly.


This doesn't use the RGBmax trick (the blue and green vertices are less than full brightness), but still has ignored the effect of display gamma.	This image is technically correct, has the correct gamma, and is also free of the nasty bright-Y artifact. It is a bit dull however, as no alternative enhancement has been made to get the maximum brightness from the green and blue corners of the triangle.