Abstract
Color appearance in anomalous trichromats may be surprisingly similar to that of color normal (CN) observers, despite the presence of an anomalous L- or M-photopigment. One possible explanation is renormalization of the spectrally-opponent mechanisms that contribute to color appearance (see: Webster, et al. 2010; Boehm, et al., 2014). Here, we explore the normalization process by studying unique yellow (UY) settings of deuteranomalous trichromats (DAs).
Three DAs (match-ranges=3, 9, and 15 Nagel-I units) and 11 CN observers set UYs by adjusting 2-deg, 18 cd/m^2 colored disks flashed at 0.5 Hz within a steady 12 cd/m^2 gray surround. The target colors fell on the maximum color circle, centered near white, of a wide-color-gamut color monitor (Eizo CG276). All observers also made UY settings by mixing 540nm and 670nm primaries in the Nagel anomaloscope.
On the monitor, DAs' UY settings were tightly clustered and fell **within** the range of CN settings, but DAs' anomaloscope UY settings had more **green** than those of CNs. Also, Rubin’s (1961) 32 DAs adjusted monochromatic lights to **redder** settings (583nm) than CNs (577nm). All these seemingly disparate results are well accounted for quantitatively by one-time rescaling of L/ M’ cone weights of Ingling & Tsou’s (1975) color-normal r-g color-opponent spectral sensitivity function, then re-balancing the color-opponent sub-stages to produce a null response for an equal-energy spectrum.
We have extended our empirical studies to other special colors viewed within gray surrounds: unique red, green, and blue, plus four “balanced” binary colors (e.g., “an orange that is equally reddish and yellowish”, cyan, purple, and lime). All but one single DA color setting fell within the CN setting-ranges.
These results suggest that, despite the presence of an anomalous M-photopigment, renormalization produces a perceptually-scaled DA hue space that closely approximates CN space (see Bonnardel, 2006; Boehm, et al., 2014 for related work).