Find in Library

The temporal protocols used in stereoscopic 3D (S3D) displays have many parameters that affect the visibility of temporal artifacts. Flicker is visible at low presentation rates. Motion artifacts—judder, edge banding, and edge blur—are salient at low capture rates. Different stereoscopic techniques can introduce distortions in apparent depth. For example, temporal interlacing (i.e., alternating presentation to the two eyes) can cause depth distortions for objects moving horizontally across the screen. Spatial interlacing (i.e., presenting information to the two eyes simultaneously, but with alternate pixel rows going to different eyes) avoids depth distortions, but causes a reduction in effect spatial resolution. We performed a psychophysical experiment to measure perceptual thresholds for three temporal artifacts in S3D displays: flicker, motion artifacts, and depth distortions. We used a stereoscope consisting of two CRTs running at 200 Hz. A high-precision synchronizer drove the two CRTs thereby allowing us to simulate a wide variety of S3D display protocols accurately. We measured the visibility of the three artifacts while varying presentation rate, capture rate, object contrast, and object velocity. We tested both temporal- and spatial-interlacing protocols. From the results, we developed a perceptual model of artifact visibility in S3D displays.