Abstract
The visual system is comprised of several binocular subsystems, at least two of which compute estimates of absolute and relative disparities. Absolute disparities are extracted at early stages of processing, whilst relative disparities may be computed by differencing the outputs of two or more absolute disparity detectors. A prediction of a simple differencing model is that the contrast dependence of absolute and relative disparity coding should be the same. We tested this prediction by recording steady-state visual evoked potentials (N=7) to stimuli designed to isolate disparity and by measuring response components associated with absolute vs relative disparity processing. Stimuli were dynamic random-dot stereograms alternating at 2Hz between a disparate surface and a zero-disparity plane. Disparity amplitude was swept, increasing the depth modulation over the course of each 10-second trial. Stimuli were shown at 8 log-spaced contrast levels ranging from 2.5-80% Michelson contrast. Disparate stimuli either contained only absolute disparities (a flat plane with no references) or relative disparities (a corrugated surface at 0.75 cpd). We replicated earlier results showing that the sustained 1f1 response best captures relative disparity processing, whilst the transient 2f1 response is largely driven by absolute disparities, suggesting a division of labour across different temporal channels. The 1f1 response showed a steeper contrast gain, implying that the sustained, relative disparity channel is more dependent on stimulus contrast. The absolute disparity response could be driven by lower contrast levels: At low stimulus contrasts, the absolute disparity signal was greater, with relative disparity catching up and overtaking in amplitude from about 10% contrast. Our results suggest differences in the contrast dependency of absolute/transient and relative/sustained disparity mechanisms that are inconsistent with a simple differencing computation. The differential sensitivity of these two disparity sub-systems may have implications for the development of disparity mechanisms, as well as clinical interventions in amblyopia.