Implement support for "Optimal Colour Stimuli" computation.

[KelSolaar]

We are currently using pre-computed Optimal Colour Stimuli tables from MacAdam (1935) and they are not very precise, only account from Luminance domain [10, 95] with the negative side-effect of labelling a lot of Munsell Real colours outside the MacAdam limits:

import colour

xyY_r = []
for _HVC, xyY in colour.MUNSELL_COLOURS_REAL:
    xyY_r.append(xyY * np.array([1.0, 1.0, 1.0 / 100.0]))
xyY_r = np.array(xyY_r)

iml = colour.is_within_macadam_limits(xyY_r, 'C')
print(len(xyY_r))
print(len(iml[iml == False]))
# 2734
# 829

We should improve this behaviour by implementing an analytical computation for Optimal Colour Stimuli.

References

• Francisco Martínez-Verdú, Esther Perales, Elisabet Chorro, Dolores de Fez, Valentín Viqueira, and Eduardo Gilabert, "Computation and visualization of the MacAdam limits for any lightness, hue angle, and light source," J. Opt. Soc. Am. A 24, 1501-1515 (2007)
• https://webserver2.tecgraf.puc-rio.br/~mgattass/fcg/Color/aic_1MacAdamLimits.pdf
• https://github.com/rodrimc/ColourGamut
• https://www.researchgate.net/publication/44673524_Fast_and_accurate_model_for_optimal_color_computation

[gutenzwerg]

I wrote a program to calculate MacAdam-Limits. I presume it is, what was wished in this issue I did already a pull request for this Feature/macadam limits #768

Every single whavelenght shoud be accurate enough. If necessary, it is possible to modify the program to use different cmfs instead of only CIE-31 2° by default.

[gutenzwerg]

As far as I see. My function "macadam_limits" is not build in in the actual version. Am I right? I need help in resolving the moand conflicts written at the pull request. It all relates to stuff of rights and features within comment-areas. I have no idea what to write in. Can anyone help me to resolve the conflicts? Otherwise this feature can not be implemented.