Abstract
After many weeks of practice, subjects can report the number of clearly-separable transparent depth planes from 2 to 6, in briefly-presented (400ms) stereo displays of 96 randomly-located x’s. They report the number from 7 to 10 with decreasing but still above chance accuracy (VSS 2010). Jittering the disparity differences between planes and jittering the number of elements in each plane, to remove spurious cues, reduced the range by only 1, so we have concluded that subjects can learn to see 6 or more transparent depth planes in parallel (i.e., without scanning or vergence movements across the displays). But how organized is the information in these depth planes? We have one finding. When given a constant number (6) of planes, subjects can enumerate how many of them contain +’s when the rest contain x’s, and subjects can report which particular plane, from 1 through 6, contains +’s, when all the rest contain x’s. We conclude that experienced subjects can learn to see, in one glimpse, a ‘depth frame’ comprising transparent depth planes with known features in a known order.
Meeting abstract presented at VSS 2015