This is the 25th in a series of posts on color reproduction. The series starts here.
I’ve cleaned up the u’v’ plot from yesterday a bit, and have done CIELab and CIELuv chromaticity representations of the three-way camera/profile/raw developer comparisons.
Here they are:
I added the color shadings to the Lab and Luv plots. They were computed at L* = 60, and the colors are not strictly accurate, partly because of the gamut of sRGB, and partly because of the Lab and Luv gamuts at that luminance.
One thing is interesting to note. If these were perfect color spaces for gamut mapping and calculating chromaticity shifts, apparent hue would not change as you mover radially at a constant angle from the origin. That is not the case in all parts of either Lab or Luv, as you can see by closely inspecting the two plots above.
It is especially not the case in the magenta/blue transition of Lab. Note that saturated blues near 5:30 or 6:00 on the plot become magentas as you move radially towards the origin. This is the origin of the — formerly dreaded by color scientists — “purple sky” problem, which caused skies to shift in a really ugly direction as the gamut of the printer in the blue/magenta decreased.