Abstract
To examine influences of selective spatial attention on stereo performance we measured disparity thresholds to detect dual targets presented simultaneously in the central and in the peripheral visual field, respectively. A random dot stereogram subtending 29° by 29° was presented for 0.2 s while the observer fixated at the central fixation point. A square-shaped target was defined by crossed disparity. We used the following four experimental conditions. The first two provided the baseline performance. In the center-only condition the target was of 1° size and presented at 2° eccentricity. The direction of the target relative to the fixation point was either one of eight directions in 45° steps. The observer's task was to indicate the direction of the target in 8AFC manner. In the surround-only condition the stimulus pattern was the same except that the size and eccentricity of the target were five times larger than those in the center-only condition. In the center-priority condition the central and peripheral targets were presented simultaneously while the observer paid more attention to the center target. In the surround-priority condition more attention was paid to the surround target. For comparison thresholds to detect luminance-defined targets were determined using the same experimental paradigm. The results showed that threshold elevation occurred almost exclusively for the less attended targets, and that the magnitudes of threshold elevation were in up to 0.5 log-unit range for most observers. They were much larger than those for luminance detection. It appears that precise stereoscopic depth judgments need spatial attention.