Abstract
In order to perceive motion accurately, an observer must perceive both speed and direction. We investigated whether speed and direction are processed separately or together by measuring the perceived speed and speed discrimination thresholds for global flow stimuli (e.g., Watamaniuk, Sekuler & Williams, 1989). If speed and direction are processed together, then increasing the range of directions in the global flow stimulus should impact perceived speed and speed discrimination thresholds.
Observers viewed random dot cinematograms in which each dot moved with a constant step size from frame-to-frame and randomly chose its direction each frame from a distribution of directions (frame rate = 60 Hz). To measure perceived speed, observers were presented with two stimuli in each trial where the standard contained dots that moved in a single direction (upwards) and the comparison contained dots that moved in a range of directions spanning 90° or 120° with a mean direction of upwards. The speeds of the standard were 6, 9, 12, and 15 deg/sec. The speed of the comparison varied from trial to trial according to a one-up, one-down staircase that converged on the point of subjective equality. To measure speed discrimination, the method of constant stimuli was used and observers were presented with a single stimulus on each trial and asked to judge if the stimulus moved faster or slower than the implicit mean speed. Thresholds were measured for four mean speeds, 6, 9, 12, and 15 deg/sec, and for stimuli spanning three different direction ranges, 0°(all dots moved in a single direction), 120° and 360°.
Results showed that perceived speed of global flow decreased as direction bandwidth increased and that speed discrimination was poorer (higher thresholds) when direction noise was added to the stimulus. Taken together, these results lend support to the idea that speed and direction are processed together.