Abstract
Perception of a stable visual environment depends on mechanisms that monitor the agreement between visual and non-visual cues to head movement. Visual cues include a combination of optic flow and oculomotor signals needed to bring retinal motion into head coordinates. Non-visual cues include vestibular signals as well as efference copies of motor commands to rotate the head on the body. Here we investigate specifically how characteristics of the visual scene impact observers’ ability to detect visual-vestibular conflict. Experiments were conducted using a head-mounted display (HTC Vive Pro Eye) with fixation behavior monitored by its embedded eye tracker, and head-movement was captured using an external motion capture system (Optitrack Prime 13). On each trial, participants made active yaw head movements of ~30 deg over 1.5 sec while fixating a scene- or head-fixed point. During the head movement, the gain on the visual scene motion was manipulated. Participants were asked to report whether the gain was too low or too high, that is, if the environment appeared to be moving with or against head movement, respectively. Fitting a psychometric function to the resulting data yields the gain perceived as stationary (PSE) and the range of gains that are compatible with perception of a stationary visual environment (JND), referred to by Wallach as the Range of Immobility. Participants were tested using a virtually rendered optokinetic drum with either a low or high-frequency stripe pattern. Our results show lower visual gain (PSEs) in the high-frequency condition suggesting greater perceived speed of scene motion consistent with the known effect of spatial frequency on perceived speed of retinal image motion. Sensitivity to conflict was also reduced (higher JND) in the high-frequency condition suggesting increased signal-dependent noise on estimated scene motion.