Abstract
This study compares stereo-slant discrimination thresholds of vertically separated random dot surfaces viewed either with or without saccadic gaze shifts. How is the stereo-slant perception affected by retinal eccentricity? Do eye movements lower thresholds even though they produce vergence fluctuations that introduce additional noise?
Stereo-slant discrimination was tested using random dot patterns (V 1.5 × H 8 deg). Yaw-slant was produced by horizontal magnification of one ocular image. Reference and test patches were presented for 167 ms each, either simultaneously or sequentially (ISI = 400 ms). The reference patch was centered at the fixation point and the test patch was presented at retinal eccentricities ranging from 1 to 16 deg above the reference. Condition 1: simultaneous targets were viewed with fixation maintained on the reference surface. Condition 2: sequential targets were also viewed with steady fixation. Condition 3: gaze was shifted between sequentially presented targets.
In conditions 1 and 2, thresholds were lowest for the smallest target separation, and remained constant over a small range of target separations. The threshold was lower for the simultaneous than sequential stimulus presentation. At larger target separations, threshold increased proportionally with the square of the retinal eccentricity. The knee where the constant and rising portions of the function meet occurs at a smaller retinal eccentricity for the simultaneous than sequential condition. This knee describes the retinal eccentricity at which position uncertainty exceeds the noise sources that determines threshold at the fovea. The function for condition 3 was flat and similar to the foveal threshold for the sequential condition 2. It indicates that noise introduced by vergence fluctuations associated with saccadic gaze shifts is smaller in comparison to the other noise sources limiting threshold.
Supported by NEI grant EYO 8882