Reanalysis of lambda max variations in the Stiles-Burch 10 degrees color-matching functions.

Individual differences in the color matches made by normal observers can be attributed in part to small interobserver variations in the spectral peaks (lambda max) of the cone sensitivities. I compared two different analyses of these lambda max variations that were both based on the Stiles-Burch 10 degrees color-matching functions [Opt. Acta 6, 1 (1959)]: one that suggested that the lambda max values for individual cone classes fall into discrete subgroups [J. Neitz and G. H. Jacobs, in Colour Vision Deficiencies IX, B. Drum and G. Verriest, eds. (Kluwer Academic, Dordrecht, The Netherlands, 1989)] and one that failed to find discrete clustering [J. Opt. Soc. Am. A 5, 1722 (1988)]. I conclude that there is not strong evidence for discrete lambda max variations in the Stiles-Burch matches.