Abstract
Early attempts at estimating MAE magnitude used matching paradigms to measure perceived speed of static test stimuli. Because of the difficulties in matching a stimulus that changes position with a static test stimulus, this approach was replaced with duration as a measure of MAE magnitude. However, when a static post-adaptation test stimulus is replaced with dynamic noise and the matching stimulus is composed of a mixture of noise and signal dots (with a biased motion direction), observers cannot distinguish between the two (Hiris & Blake, 1992). This characteristic of dynamic noise stimuli, in conjunction with our recent findings that perceived speed of random dot patterns varies with their signal-to-noise ratio, offers a means to measure MAE speed accurately. Observers fixated a central spot. During adaptation a dot pattern, moving upwards, was presented within an aperture to the left of fixation; a dynamic noise pattern (zero motion coherence) was presented to the right. The adaptation phase lasted 60 seconds. Thereafter observers were presented with alternating test and ‘top up’ phases. Each top up phase, lasting 10 seconds, was used to maintain adaptation. During each test phase a dynamic test stimulus, with no biased motion direction, was presented in the adapted field. The matching stimulus, which contained biased upwards motion, was presented in the non-adapted field. The observers' task was to judge which stimulus was moving fastest. The speed of the matching stimulus was varied across trials. Observers were tested with a range of adaptation speeds (1 – 12 deg/sec), with at least 12 hours between each speed condition. The perceived speed of the MAE as a function of adaptation speed described an inverted U-function, peaking at an adaptation speed of approximately 8 – 10 deg/sec. These results concur with earlier MAE ‘duration’ experiments (Ashida & Osaka, 1995). We are currently investigating whether the effect is speed or temporal-frequency tuned.