Purpose. We evaluate the orientation selectivity of red-green and blue-yellow chromatic mechanisms using an external noise paradigm that allows the assessment of the internal orientation noise, the relative sampling efficiency, and the orientation bandwidth of the underlying orientation-tuned mechanisms.

Methods. The task required the measurement of orientation acuity (detection of orientation change) in a temporal 2AFC staircase method. Stimuli were patches of orientation noise defined in the Fourier domain multiplied by a Gaussian envelope in the space-time domain (sigma_x = 1 deg, sigma_t = 500 ms). Orientation acuity (sigma_o) was measured as a function of peak frequency, spatial bandwidth, and stimulus bandwidth in orientation (sigma_e). Internal orientation noise (sigma_i), relative sampling efficiency (N), and orientation bandwidth sigma_e(knee) of the underlying mechanism were derived by fitting the data with a noise model:

sigma_o = sqrt(sigma_i^2+sigma_e^2/N)

and

sigma_e(knee) = sqrt(N).sigma_i

Stimuli were cardinal, isolating each of the three postreceptoral mechanisms, and matched in multiples of detection threshold.

Results. We find that orientation bandwidth and internal orientation noise are significantly greater in the chromatic than the achromatic systems. Preliminary results indicate that red-green orientation selectivity depends on the spatial properties of the stimulus (peak frequency and spatial bandwidth).

Conclusions. We conclude that color vision (red-green and blue-yellow) has a moderate deficiency in orientation selectivity. This may account for the small differences we have found between color and luminance vision on contour integration and shape discrimination tasks (Mullen et al, Vis. Res. 40, 2000; Mullen & Beaudot, Vis. Res., 2002).

Supported by CIHR Grant MOP-10819