Abstract
In order to determine the chromatic contrast sensitivity function, the luminance equilibrium of chromatic pairs should be evaluated with the actual visualization display. This evaluation should be done at different eccentricities when the visual stimulations are DOGs (Difference Of Gaussians), because the spatial frequency variation modifies the spatial extent of the stimulation.
The luminance equilibrium evaluation method of chromatic pairs was based on the apparent motion paradigm (Anstis, Cavanagh 1983). We filled an annulus with two-colour radial sectors. When the colours differed in luminance, the stimulation looked like the wings of a windmill (Anstis, S. 2000). At equiluminance, the motion appeared to vanish.
The experimental method was a constant stimuli method with ten values of luminance for one colour and eleven repetitions for each level. A two alternative forced-choice procedure (clockwise or counterwise rotation) was used. Four colour pairs were tested: red-green, yellow-blue (colour-opponent pairs), cyan-magenta, yellow-purple (cone-opponent pairs). Two eccentricities were studied: for each annulus, the inner and outer annulus diameters being respectively 0,4°–2,4° and 2,8°–5°. Tests were displayed on a flat CRT screen with a mean luminance of 16 cd.m-2 for the yellow-blue pair and 30 cd.m-2 for the others. Seeven colour-normal observers took part in the experiment.
The results show that the luminance equilibrium is identical for the two sizes of annulus except for the yellow-blue pair. For this pair, the mean ratio: yellow luminance/blue luminance is equal to 1 for the smallest eccentricity and decrease to 0.8 for the largest.