Color, contrast sensitivity, and the cone mosaic.

Abstract

This paper evaluates the role of various stages in the human visual system in the detection of spatial patterns. Contrast sensitivity measurements were made for interference fringe stimuli in three directions in color space with a psychophysical technique that avoided blurring by the eye's optics including chromatic aberration. These measurements were compared with the performance of an ideal observer that incorporated optical factors, such as photon catch in the cone mosaic, that influence the detection of interference fringes. The comparison of human and ideal observer performance showed that neural factors influence the shape as well as the height of the foveal contrast sensitivity function for all color directions, including those that involve luminance modulation. Furthermore, when optical factors are taken into account, the neural visual system has the same contrast sensitivity for isoluminant stimuli seen by the middle-wavelength-sensitive (M) and longwavelength-sensitive (L) cones and isoluminant stimuli seen by the short:wavelength-sensitive (S) cones. Though the cone submosaics that feed these chromatic mechanisms have very different spatial properties, the later neural stages apparently have similar spatial properties. Finally, we review the evidence that cone sampling can produce aliasing distortion for gratings with spatial frequencies exceeding the resolution limit. Aliasing can be observed with gratings modulated in any of the three directions in color space we used. We discuss mechanisms that prevent aliasing in most ordinary viewing conditions.There has been considerable recent interest in understanding how color and luminance information are processed in the visual system. It is well known that the mechanisms sensitive to chromatic modulation have inferior resolution to those sensitive to luminance modulation (1), but the reasons why this is so are not yet clear. This paper examines the role of the first few stages of the visual system in detecting spatial patterns that vary in chromaticity or luminance. Wilson Geisler (2) has developed an elegant computational approach to this problem based on the theory of ideal observers. We describe an ideal observer similar to Geisler's and use it to examine the effect on contrast sensitivity of optical factors such as those associated with the cone mosaic. We then compare ideal observer performance with new measurements of human contrast sensitivity for patterns modulated in different color directions. This comparison reveals the properties of the neural mechanisms that underlie human performance. This paper summarizes work described by Sekiguchi et al. (3,4) to which readers are referred for details of the psychophysical experiments and computational model. Stimuli for Measuring Ideal and Human Contrast SensitivityWe computed the contrast sensitivity of the ideal observer and measured the contrast sensitivity of human observers with the same stimulus set. For the ideal observer, analytic descriptions of the stimuli were used as input to a computa...

Senescence of spatial chromatic contrast sensitivity I Detection under conditions controlling for optical factors

“…Because of relatively sparse retinal sampling, the highest spatial frequencies used in this experiment approach the upper resolution limit of the S-cone system. 42 Finally, luminance artifacts due to chromatic aberration or imperfect luminance equations with the HFP technique could cause such a convergence of sensitivity curves. Despite our best efforts to control luminance artifacts, it is possible that luminance intrusion occurred and biased sensitivities toward higher values for some of the older observers.…”

Detection of Fruit and the Selection of Primate Visual Pigments for Color Vision

“…In trichromatic primates, the nervous system combines M and L cone signals to give a luminance signal, which is used for "color-blind" tasks such as motion detection and shape recognition (Livingstone and Hubel 1988;Mollon 1989). Differing spectral inputs could corrupt this luminance signal, much as TV signals are corrupted when fine patterns such as striped clothing are rendered as shimmering colored moiré patterns (Williams et al 1993). This problem may result in the spectral separation of the M and L pigments being limited to a value below the optimum for color vision ( fig.…”

Color—luminance relationships and the McCollough effect

“…Similar results were also obtained for gratings of 3 or 12 cycles/deg, when the display was viewed directly at the closer distance (Figure 8). The higher frequencies should be near (12 cycles/deg) or well above (24 cycles/deg) the resolution limit of the S cones (Green, 1968;Stiles, 1949;Williams & Collier, 1983;Williams et al, 1993), yet MEs for these patterns were readily visible and varied with frequency in a similar way for the L-M and S-(L+M) axes.…”

“…For example, Scones contribute little to conventional measures of luminance (Lennie, Pokorny, & Smith, 1993). Moreover, because there are few S cones and they are sparsely distributed (Curcio et aI., 1991;De Monasterio, McCrane, Newlander, & Schein, 1985;Marc & Sperling, 1977), the spatial resolution limit ofthe S cone mosaic is substantially lower than the L or M cone mosaic (Green, 1968;Stiles, 1949;Stromeyer, Kranda, & Sternheim, 1978;Williams & Collier, 1983;Williams, Sekiguchi, & Brainard, 1993). We therefore asked whether these differences in spatial sensitivity would result in differences in the spatial dependence of the aftereffect.…”