Commercial virtual reality displays: Issues of performance and simulator sickness from exocentric depth-perception tasks

Abstract

Issues in both depth-perception performance and simulator-sickness effects may hinder future applications of virtual reality. Applications for depth perception are varied, including military, sports, and architecture domains. The purpose of this study is to compare different commercial-off-the-shelf head-mounted displays and their content, through depth-perception tasks. We also investigate simulator sickness profiles to promote further risk mitigation in such displays. Specifically, the HTC Vive and Oculus Rift Consumer Version 1 were examined in between-groups and within-groups designs, with both head-mounted displays supplying the same exocentric depth-perception tasks across three virtual environments of increasing ecological complexity. Depth-perception performance was measured objectively and simulator sickness was measured subjectively via questionnaire. Results, analyzed using non-parametric statistics, indicated significant differences for performance and simulator sickness, for both between- and within-groups studies. Although distance changes were detected equally between displays, a tradeoff occurred where the Vive had better distance classification, but the Rift had shorter time duration. Hardware issues are considered for performance differences. For post-test simulator sickness, Vive and Rift only differed in terms of fatigue, an oculomotor-based issue with probable user-avoidance strategies; the difference may relate to latency. Performance within Rift and Vive only partially showed a benefit for high-realism scenarios, yet both displays provide distinct scenario-change baselines for personnel selection. Simulator sickness for both devices showed Oculomotor subscale increases in pre-to-post test questionnaire scores, with the Vive having an additional increase in Total scores. The latter Oculomotor increase may be driven by blur. In respect to the observations, multiple factors exist for selection, including performance and simulator sickness tradeoffs.