Abstract
Purpose: We have previously demonstrated that motion discrimination for red-green chromatic stimuli is based on luminance signals under certain conditions, specifically those of linear motion (Yoshizawa et al., Vis.Res. 2000), signed apparent motion (Yoshizawa et al., Vis.Res. 2003), and grating motion (Mullen et al. Vis Res. 2003). Here we test whether the motion of blue-yellow chromatic stimuli is similarly contaminated by luminance signals or is genuinely mediated by chromatic mechanisms.
Methods: Stimuli were gabor patches (spatial frequency = 1.5cpd, σ = 0.66 degrees) calibrated to isolate either the S-cone or L/M cone opponent pathways. We investigated the masking effects of superimposed luminance noise (1-dimensional, temporally dynamic) on contrast thresholds for stimulus detection and direction discrimination using a temporal 2AFC staircase method. Thresholds were measured for stimulus drift rates of 0.75 – 6.0 Hz. Individual isoluminant points were determined using a minimum motion technique.
Results: We find that for S-cone opponent stimuli neither detection nor direction discrimination thresholds were masked by luminance noise, and a large gap was maintained between detection and direction discrimination thresholds. For L/M cone opponent stimuli, detection thresholds were not affected by luminance noise but direction discrimination was masked by increasing luminance noise contrast.
Conclusions: Unlike their L/M counter parts, direction discrimination thresholds of S-cone opponent gabors are not based on luminance signals, and appear to be genuinely chromatic. Hence we conclude that temporal luminance artifacts, previously termed ‘temporal chromatic aberration’, are specific to L/M cone systems.
NSERC Grant RGPIN 183625-05