Purchase this article with an account.
L. N. McLin, Jr., J. R. Dykes; Color Naming Across the Mesopic Range as a Function of Photoreceptor Sensitivity. Invest. Ophthalmol. Vis. Sci. 2010;51(13):6291.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Classic research has focused on changes in color appearance across the photopic range and the lack of chromaticity in the scotopic range. This research investigated changes in color appearance in the mesopic range as a function of specific photoreceptors.
Across two days, 24 observers used an extension of the Levele 2 Universal Color Language to name the colors of 48 stimuli on a Sony GDM-C520 CRT viewed through neutral density filters that reduced the stimulus luminance from 17.5 cd/m2 to 0.0011 cd/m2 in eight 0.6 OD steps. The same task was also performed viewing through a neutral density goggle and through a colored goggle with average transmittance of 27%. Half of the observers used an orange goggle (with a dominant l-cone transmittance) and half a blue goggle (with a dominant s-cone and rod transmittance). The stimuli were similar in photopic luminance and sampled the entire gamut. There were 16 stimuli that matched Macbeth ColorChecker stimuli. Four stimuli were anchors: the 3 vertices of the gamut (except for blue) and Illuminant C. The other 28 stimuli evenly sampled across the gamut in CIE u’v’space.
The color name responses were summarized into nine chromatic and three achromatic categories. The probabilities of those responses given for each stimulus were used to create a radial color zone space based on hue angles and chromatic radii. With no filter, the color zone space was fairly stable down to 1.1 cd/m2. As expected, borders shifted/disappeared and chromaticity substantially decreased below 0.28 cd/m2. The data with the 27% neutral goggles replicated those changes after adjusting for the reduced luminance. The color zones were severely disrupted at 6 cd/m2 using the orange goggles (which blocked s-cone wavelengths). While the borders were shifted, the disruption of the color zones while using the blue goggles (which blocked the l-cone wavelengths) were only modest at 1cd/m2.
The changes in color naming zones across the mesopic range demonstrated the importance of the s-cones and the blue-yellow opponent color channel. Clearly the s-cone is much more important than the l-cone for color naming in the mesopic range.
This PDF is available to Subscribers Only