Self-motion and Presence in the Perceptual Optimization of a Multisensory Virtual Reality Environment

Sammanfattning: Determining the perceptually optimal resolution of multisensory renderingmight help to foster the development of cost-effective, highly immersivemulti-modal displays for mediated environments (e.g. virtual and augmentedreality). The required sensory depth of stimulation can be quantified usinghuman centered methodologies where end user experiences serve as a basis foruni- and cross-modal optimization of the sensory inputs. In the psychophysicalstudies presented in this thesis, self-reported presence and illusoryself-motion (vection) indicated salience of auditory and multisensory cues indesign of perceptually optimized motion simulators.Contribution of auditory cues to illusory self-motion has been largelyneglected until very recently and papers A and B present studies on purelyauditory induced vection (AIV). Paper A shows that rotating auditory scenessynthesized using individualized Head-Related Transfer Functions (HRTFs) aremore instrumental for presence compared to generic binaural synthesis. Study ontranslational AIV in paper B shows that inconsistent auditory scene mightsignificantly decrease self-motion responses. Paper C and D demonstrate thatbi-sensory stimulations increase presence and self-motion ratings as expected.In paper C additional vibrotactile stimulation increased translational AIV andpresence ratings, especially for the stimuli containing the auditory-tactileengine metaphor. Paper D extended paper A results for rotational AIV showingthat spatial resolution of rotating auditory scenes can be greatly reduced whencombined with visual input.This thesis shows that sound plays important role in the illusory self-motionperception and it should be carefully used in multi-modal motion simulators. Thepresented findings suggest that a minimum set of acoustic cues can be sufficientfor eliciting a self-motion sensation, especially if other modalities areinvolved. However, perceptual consistency of the created auditory and multimodalscenes should be assured in the design of the next generation of motionsimulators.