Contrast dependence and mechanisms of masking interactions among chromatic and luminance gratings.

The contrast dependence of simultaneous masking has been measured using isochromatic yellow-black luminance sinusoids and isoluminant red-green chrominance gratings. Masking functions for all four combinations of chromatic and luminance masks and tests are reported. In the two same-on-same conditions (luminance mask/luminance test and chromatic mask/chromatic test) these functions (increment threshold contrast versus mask contrast) have the typical dipper shape and are almost identical when test and mask contrasts are normalized to the unmasked contrast thresholds. The contrast dependence of the luminance mask/color test and color mask/luminance test functions are quite different. The luminance mask/color test shows facilitation over a broad range of both subthreshold and suprathreshold contrasts of the luminance mask. In the color mask/luminance test condition facilitation is never observed, but at suprathreshold contrasts a 2-cycle/degree (c/deg) chromatic grating masks a 2-c/deg luminance grating as strongly as does a luminance mask. The luminance mask/chromatic test results are invariant over the 0.25-2-c/deg spatial-frequency range, whereas the robust masking of luminance by color at 2 c/deg diminishes at lower spatial frequencies. The spatial-frequency selectivity of the luminance-facilitates-color interaction is much broader than facilitatory interactions in either the color-color or luminance-luminance conditions. Possible mechanisms of color-luminance interactions are considered. The lack of facilitation in the color mask/luminance test condition precludes a simple pedestal interpretation of this masking interaction. The data are, however, consistent with models that invoke inhibitory or more elaborate excitatory masking interactions.