Abstract
The transduction mechanisms in the visual system are nonlinear. In pedestal masking the detection of contrast increments is facilitated at low pedestal contrasts and masked at high pedestal contrasts. Hence increment thresholds measured over a range of pedestal contrasts form a "dipper" function whose shape reflects the underlying contrast response function (CRF). Here, we exploit this relationship to measure the CRFs for achromatic and isoluminant red-green (RG) color contrast. Contrast increment thresholds for Gabor patterns (0.375 c/deg, 2 Hz) with either achromatic or isoluminant RG (L/M cone opponent) contrast were measured for three observers in a pedestal masking paradigm. CRFs were fitted to the individual trial responses using maximum-likelihood fitting: response = (contrast^p)/(z+contrast^q). Although cone contrast sensitivity was on average 4.91±0.62 times higher for the isoluminant RG stimuli, the shapes of the dipper functions for the two conditions were similar. The fitted transducer parameters (p and q) were found to be similar across the achromatic and RG chromatic CRFs. Between three subjects, average p values of 2.36±0.47 and 2.38±0.41 were found for achromatic and chromatic CRFs, respectively. The respective q values were 1.91±0.41 and 1.87±0.30. Regardless of individual sensitivity differences, these similarities between achromatic and color perception are consistent within each subject. We find that the shapes of the CRFs for achromatic and RG color contrast are very similar. This is important for studies that normalise the contrast responses to achromatic and color stimuli by scaling them in multiples of detection threshold. Differences in the shapes of the CRFs would make this normalisation method invalid. Future CRF measurements for achromatic stimuli at higher temporal rates (0.375 c/deg, 10 Hz) will be used to compare the supposed "magnocellular" CRF against the "parvocellular" CRFs measured here.
Meeting abstract presented at VSS 2017