Perceiving scene layout through an aperture during visually simulated self-motion.

This study tested scene perception during depicted self-movement. In Experiment 1, viewers reproduced angular scene configurations, revealed over time by camera motions in depth. Viewers used visible flow to update off-screen locations of landmarks that had been drawn from view and thereby perceived scenes in depth beyond the display's boundaries. Experiments 2 and 3 tested whether offscreen space is perceived as a function of depicted velocity and time and whether information from prior views is used. Viewers predicted when lateral movement would reveal a peripheral landmark whose location was shown in a prior panoramic view. Accurate responses, observed under certain conditions, would result if viewers perceived space according to S = integral of V dt (D. Algom & L. Cohen-Raz, 1984, 1987; R. Jagacinski, W. Johnson, & R. Miller, 1983). When required to picture wide spans in space yet to emerge, however, viewers responded as if these spans were compressed. The theoretical implications of the ability to retrieve, transform, and apply information after a discontinuous transition from a prior view are discussed.