Self-characterstics and sound in immersive virtual reality — Estimating avatar weight from footstep sounds

This experiment aimed to investigate whether a user controlling a full body avatar via real time motion tracking in an immersive virtual reality setup, would estimate the weight of the virtual avatar differently if the footstep sounds are manipulated using three different audio filter settings. The visual appearance of the avatar was available in two sizes. The subjects performed six walks with each audio configuration active once over two ground types. After completing each walk, the participants were asked to estimate the weight of the virtual avatar and the suitability of the audio feedback. The results indicate that the filters amplifying the two lower center frequencies altered the subjects estimates of the weight of the avatar body towards being heavier than when compared with the filter with the higher center frequency. There were no significant differences between the weight estimates of the two groups using the different avatar bodies.