Abstract
The percent coherence threshold in a random-dot kinematogram (RDK) global motion task depends on contrast and dot displacement. Does threshold also depend on whether signal and noise dots are presented to same or different eyes? We compared performance under three binocular viewing conditions in a 2AFC net motion discrimination task (up vs. down). In a 7 degree circular aperture at a 114 cm viewing distance, coherently moving signal dots and randomly moving noise dots were displayed to either eye entirely ( “monocular” ), or to the two eyes respectively ( “dichoptic” ), or mixed among the two eyes equally ( “binocular” ). Threshold percent coherence was estimated from a 3-down-1-up staircase procedure and converted to signal-to-noise ratio (SNR). Dot lifetime was strictly 2 frames at 30 Hz (2-frame motion), and the stimulus duration was 300 ms. Trials from different viewing conditions were randomly intermingled. We tested at two contrast levels (33%, 6%) and three displacements (0.04, 0.11, and 0.27 degree/frame) for a total of 3x2x3 = 18 viewing conditions. All 7 subjects had normal binocular vision. Five of our subjects replicated Seitz, Pilly, and Pack (2005), who reported that high luminance contrast was helpful when the displacement of dots was large, but harmful when small; however, two of our subjects showed little reduction in their high-contrast advantage even at the smallest displacement. Remarkably, dichoptic viewing was significantly better than monocular or binocular viewing at high contrast, and significantly worse at low contrast; this interaction was most dramatic for large displacements. We conclude that global motion mechanisms are variable within the population and that different pathways may operate to combine motion signals across the eyes depending on contrast.
Meeting abstract presented at VSS 2015