Abstract
It has been shown that distance perception, as measured by subjects walking to previously viewed targets, is compressed in virtual reality head-mounted displays (HMDs) relative to normal viewing. This disparity between normal viewing and HMD viewing has been investigated in a number of studies. Thompson et al. showed that the level of detail in a virtual environment is not a significant factor in the amount of distance compression experienced (Thompson et al., 2003). Creem-Regehr et al. showed that by themselves, neither monocular viewing nor restricted field of view caused distance compression in real-world viewing conditions (Creem-Regehr et al., 2003). These findings led us to hypothesize that the distance compression was the result of the optical aspects of the HMD system. We mounted a FOV-matched video camera system on the HMD (V8 – 60 deg. diagonal FOV), and fed the live video back to the HMD. This novel apparatus (inspired by an informal observation in Loomis & Knapp, 2003) was designed to differ from real-world viewing conditions only in terms of signal delay, physical properties (the weight of the HMD), and optical properties, specifically field-of-view, focal length, and image resolution. A direct walking task was used to compare this novel apparatus to both unrestricted monocular viewing, and monocular viewing with a restricted field of view. Distance compression in the video-feed HMD condition was significantly greater than that found in either the unrestricted monocular or field-of-view restricted monocular viewing conditions. This suggests that distance compression in HMD systems is due to the physical properties of the display system, or the optical properties of the head-mounted display, although the substantial time lag in our video-feed HMD system could also be responsible, in this case.
Hans Wallach Fellowship to RBM and Swarthmore College Faculty Research Grant to FHD