Abstract
Subjects observed a random-dot pattern that moved within an invisible aperture. In experiment 1 the pattern moved horizontally for 0.9 s, then changed its direction of motion by an angle between 12° and 180° for a time t and returned to horizontal motion. The threshold duration t for detection of the direction change was measured for speeds of motion of 2, 4, 8, and 16 deg s−1. The thresholds decreased with increasing angle of change, reaching an asymptotic level of 10 – 15 ms at about 60° for speeds between 4 and 16 deg s−1. It is suggested that the asymptotic course of this relationship reflects the bandwidth of the directional tuning of the motion-detection system. In experiment 2 we found that the temporal thresholds for detection of a velocity reversal (180° change in direction) are equal to those for detection of an increase in speed by a factor of three, and that the time to detect an abrupt cessation of motion is nearly equal to the time for detection of a change in speed by a factor of two. Measures such as ‘delta’ contrast and Michelson contrast are shown to be inappropriate for describing the detection of these changes in velocity. We suggest that the absolute value of the vector difference V2-V1 between the velocities before and after the change may be a meaningful measure for the description of the detection performance.
Get full access to this article
View all access options for this article.
