Sonic simulation in VR/AR contexts

Opera singers find it challenging to access performance venues for rehearsal. The challenges include the busy timetables of venues and limited travel windows caused by budget or pandemic restrictions. Limited rehearsal can impair a singer’s performance. Virtual reality and augmented reality technologies (VR and AR) make it possible for singers to rehearse remotely. Most VR and AR technologies focus on visual representation. Auditory representation receives less attention. The majority of self-voice auralization methods are limited in venue choice, customizability and degrees of freedom. They may also require specialised equipment and inaccessible or unaffordable rehearsal spaces. Two methods are developed in this thesis that can address these limitations. They are based on geometrical acoustics (GM) and on Ambisonics (AM). Both methods focus on the simulation of stage acoustic conditions such as those encountered at performance venues. The thesis provides an overview of these methods. Also, real-time implementation and subjective evaluation of the GM are discussed. Listener tests were conducted with nine voice students and professionals. The evaluations examine the perception of sound field rotation, personal preference for simulations and the plausibility of the simulation. The evaluations were performed inside the simulated venue so that a blind comparison could be conducted between the real and simulated acoustics.