STIMULI: A photo-realistic virtual replica of the Tübingen market place was presented via a curved projection screen (84x63° FOV). For vestibular stimulation, subjects were seated on a Stewart motion platform. TASK: Subjects were rotated consecutively to random orientations and asked to point as accurately and quickly as possible to 4 out of 22 previously-learned targets. Targets were announced consecutively via headphones and chosen to be outside of the current FOV.

Photo-realistic visual stimuli from a well-known environment including an abundance of salient landmarks allowed accurate spatial updating (mean absolute pointing error, pointing variability, and response time were 16.5°, 17.0°, and 1.19s, respectively). Moreover, those stimuli triggered spatial updating even when participants were asked to ignore turn cues and point as if not having turned, (32.9°, 27.5°, 1.67s, respectively). Removing vestibular turn cues did not alter performance significantly. This result conflicts with the prevailing opinion that vestibular cues are required for proper updating of ego-turns. We did find that spatial updating benefitted from vestibular cues when visual turn information was degraded to a mere optic flow pattern. Under all optic flow conditions, however, spatial updating was impaired and no longer obligatory. We conclude that good visual landmarks can initiate obligatory spatial updating and overcome the visuo-vestibular cue conflict.