One's own voice in auditory virtual environments

The sound of our own voices contributes significantly to our perception of real and virtual environments and has strong implications on the way we spe ak. It allows our speech to be controlled, which is of particular relevance in acoustically "difficult" situations, for example when speaking in a noi sy environment. As early as 1911 Lombard recognised that people speak loude r when they are not able to hear their own voices [1]. Since then, several investigations on the influence of a changed perception on the way we speak have been performed. However, up to now it has remained unclear how the pe rception of one's own voice affects the perception of a complete auditory s cene. Starting with a model of the different pathways relevant to the perception of one's own voice, this article describes the model's implementation in an auditory virtual environment generator. An auditory validation experiment was performed which showed that the naturalness of one's own voice is incre ased by a plausible presentation and that the system can be used as a tool for psychoacoustic research on the perception of one's own voice. In the se cond described experiment the implications of a plausible presentation of o ne's own voice on the sense of presence in virtual environments were studie d. To set up realistic communication scenarios the virtual environment gene rator was enhanced to a multi-user system so that it could for this purpose be used as an auditory teleconferencing system for two or more participant s. Applying this system communication experiments were performed and the pe rceived presence was measured. This article concludes by identifying possib le commercial applications for which a plausible presentation of one's voic e could be of relevance.