[AlanYx]

Does anyone have link to a deeper technical dive on the limitations of CIE 1931 that Apple CMF tries to address? The CIE's JTC22 (D8/D1): Optimized Colour Matching Functions for Display Colour Consistency documents seem to be behind an access wall.

From what I gather, it's an attempt to address the problem of observer metamerism, but I'm curious to know how closely it's tethered to D65.

[anikom15]

It’s independent of white point. The metamerism problem is fundamental to the methodology of the 2° XYZ colorspace because it doesn’t model the entire real response of the human eye. Instead it underestimates the response of the blue cone. This can lead to blue becoming purple as Y is reduced. This is because the experiment used lamps with colored filters (broadband light) and not lasers or LEDs (narrowband light).

It’s not just a matter of introducing a correction factor (although that exists, see Judd-Vos) because the solution is more complex than that. The practical take away is that color matching cannot just involve comparing a narrow point in isolation but also needs to have a baseline of controlled surround and specular reflections in addition to understanding the nature of the perceived color (reflective, emissive, narrowband, wideband, etc.).