Purchase this article with an account.
Mark Nawrot, Shanda Lauer, Jessica Holmin, Trevor Bartlett, Timothy Breider; Depth Constancy in the Apparently Circular Curvature Task with 3D Printed Stimuli . Journal of Vision 2016;16(12):648. doi: 10.1167/16.12.648.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
One technique used to study the subjective quale of perceived depth magnitude is the apparently circular curvature (ACC) task in which observers view a hemi-cylinder and report whether the depth (radius) is more or less than half of the height (diameter). Prior studies using computer-generated, stereoscopic, random-dot ACC stimuli have found the radius/diameter matching point changes with viewing distance, suggesting a lack of depth constancy. We used 3D printed, grey plastic hemi-cylinders to study depth constancy. Viewed on edge, hemi-cylinder width was 130 mm, diameter was 100 mm, and depth (radius) was varied between 40 and 60 mm, in 1.25 mm steps. Stimuli were front-illuminated with 120 cm of bright-white LED tape generating 1.6k lux illumination. Observers viewed the stimuli through an experimenter-controlled ferro-electric window at three distances: 53.5, 107, and 214 cm. Using a constant-stimulus paradigm with the ACC task, observers viewed stimuli in four different conditions: monocular/stationary, monocular/motion parallax, binocular/stationary, binocular/motion parallax. Psychometric functions revealed that at all distances, and all viewing conditions, the average matching radius was between 49 and 52 mm. The average standard deviations of the functions varied between 28 mm (monocular/stationary/107cm) and 2.2 mm (binocular/motion parallax/53.5cm). Using ascending and descending method of limits with both monocular/stationary and binocular/stationary viewing, average matching was between 50 and 51 mm at all three viewing distances. These results suggest accurate depth constancy with physical 3D objects and bright viewing conditions. To explore the discrepancy with previous results, a projector mounted just below the observer's line of sight projected random-dots upon the 3D figure at 214 cm viewing distance. The matching point was typically increased between 50% and 80%, similar to previous results. This suggests that previous failures of depth constancy may have been affected by a depth cue-conflict with the flat texture of random-dots comprising the ACC stimulus.
Meeting abstract presented at VSS 2016
This PDF is available to Subscribers Only