Abstract
For fixed binocular disparity (BD), the relationship between viewing distance and relative depth is well known, the distance-square law. Similarly, perceived depth from motion parallax (MP) varies with viewing distance. Here, depth of a non-fixated object (d) varies with distance to the point of fixation (f), the retinal image velocity of the object (dθ), and velocity of the pursuit eye movement required to maintain fixation (dα). The relationship, termed the motion/pursuit ratio, is: d = dθ/dα * f . Here we investigated how the cues of vergence and accommodation contribute to apparent viewing distance in MP by assessing perceived depth magnitude. Participants viewed translating random-dot MP and stationary BD stimuli at 36, 54, and 72 cm, with each stimulus type depicting a range of motion/pursuit ratios and disparities. MP stimuli were presented monocularly to the right eye and to ensure correct vergence the fixation spot of this stimulus was fused with a lone fixation spot presented to the left eye. In both conditions vergence was manipulated with a mirror stereoscope and calibrated to simulate a particular vergence angle with a dual-laser system. Accommodative load was manipulated with trial lenses. Perceived depth magnitude was indicated by the distance between the participant’s thumb and forefinger measured with a linear potentiometer. This method of perceived depth measurement was characterized by both large inter-observer differences and large intra-observer variability. Perceived depth magnitudes varied linearly with disparity, and with motion/pursuit ratios having the same pursuit velocity. Changes in vergence changed perceived depth magnitude, but less than the equivalent change in actual viewing distance. Adding a concomitant change in accommodative load did not appear to change perceived depth. Perceived depth from MP is linked to apparent viewing distance, which is not determined solely by vergence and accommodation. Furthermore, inter-digit-distance is a noisy estimator of perceived depth.
Meeting abstract presented at VSS 2015