Spatial Masking Does Not Reveal Mechanisms Selective to Combined Luminance and Red-Green Color

Detection thresholds plotted in the L and M cone-contrast plane have shown that there are two primary detection mechanisms, a red–green hue mechanism and a light–dark luminance mechanism. However, previous masking results suggest there may be additional mechanisms, responsive to combined features like bright and red or dark and green. We measured detection thresholds for a 1.2 c deg−1 sine-wave grating in the presence of a spatially matched mask grating which was either stationary, dynamically jittered or flickered. The stimuli could be set to any direction in the L,M plane. The appearance of selectivity for combined hue and luminance arose only in conditions where adding the test to the mask modified the spatial phase offset between the luminance and red–green stimulus components. Sensitivity was very high for detecting this spatial phase offset. When this extra cue was eliminated, masking contours in the L,M plane could be largely described by the classical red–green and luminance mechanisms.