Answers are grounded in this paper and cite their sources.
Calculated hue shifts could vary widely for spectrally similar light sources, which suggests a limitation of the current hue shift formulae for non-white light sources. Calculated differences in the hue appearance of 24 colour samples under 279,936 iteratively-generated test SPDs versus two reference illuminants were compared using four hue shift formulae that are based on two widely used colour spaces, CIE 1976 L* a* b* (CIELAB) and colour appearance model 2002 (CAM02). Hue shift equations did not correlate well, even when only nominally "white" light sources were considered. Results suggest that the underlying difference between the hue shift formulae might be the non-uniform scaling of redness-greenness (a) and yellowness-blueness (b) coordinates in CIELAB and CAM02-UCS. A chromatic adaptation transform correction improved the correlation between lightness in CIELAB and CAM02-UCS when the test light sources were nominally "white," but caused irregularities for non-white light sources. Further research is needed for visual assessment of the hue shifts and to evaluate the hue shift formulae.
No takes yet. Share an insight, caveat, or question.
Durmus et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: