Abstract
When we rotate our heads during head-mounted display (HMD) based virtual reality (VR), our virtual head tends to trail its true orientation (due to display lag). However, the exact differences in our virtual and physical head pose (DVP) vary throughout the movement. We recently proposed that large amplitude, time varying patterns of DVP were the primary trigger for cybersickness in active HMD VR. This study tests the DVP hypothesis by measuring the sickness, and estimating the DVP, produced by head rotations under different levels of imposed display lag (from 0 to 200 ms). On each trial, users made continuous, oscillatory head movements in either yaw, pitch or roll while seated inside a large virtual room. After, we used the level of imposed display lag for the condition, and the user’s own tracked head-motion data, to estimate their DVP time series data for each trial. Irrespective of the axis or the speed of the head movement, we found that DVP reliably predicted our participants experiences of cybersickness. Significant positive linear relationships were found between the severity of their sickness and the mean, peak and standard deviation of this DVP data. Thus, our DVP hypothesis appears to offer significant advantages over existing (general) theories of motion sickness in terms of understanding user experiences in HMD VR. Instead of merely speculating about the presence, or degree, of sensory conflict in a particular simulation, DVP can be used to estimate the conflict produced by the active HMD VR. Importantly, this DVP is an objective measure of the stimulation (not an internal model of the user’s sensory processing). Compared to its many competitors, DVP also appears to provide a simpler operational definition of the provocative stimulation for cybersickness (since it is focussed only on movements of the head; not the body or limbs).