Display size does not affect egocentric distance perception of naturalistic stimuli

Abstract

Although people are quite accurate in visually perceiving absolute egocentric distances in real environments up to 20m, they usually underestimate distances in virtual environments presented through head-mounted displays (HMDs). Several previous studies examined different potential factors, but none of these factors could convincingly explain the observed distance compressionin HMDs. In this study, we investigated the potential influence of naturalistic stimulus presentation and display size -- a factor largely overlooked in previous studies. To this end, we used an indirect blindfolded walking task to previously-seen targets. Participants viewed photos of targets located at various distances on the ground through different-sized displays (HMD, 24" monitor, and 50" screen) and walked without vision to where they thought the location of the target was. Real-world photographs were used to avoid potential artifacts of computer-graphics stimuli. Displays were positioned to provide identical fields of view (32° x 24°). Distance judgments were unexpectedly highly accurate and showed no signs of distance compression for any of the displays. Moreover, display size did not affect distance perception, and performance was virtually identical to a real world baseline, where real-world targets were viewed through 32° x 24° field of view restrictors. A careful analysis of potential underlying factors suggests that the typically-observed distance compression for HMDs might be overcome by using naturalistic real-world stimuli. This might also explain why display size did not affect distance judgments.