Perceptual Evaluation of Own Voice and Social Virtual Reality Using the MPEG-I Immersive Audio Standard

Silzle A, Disch S, Habets E, Rose S, Dennerlein F, Adami A, Lieb T. Perceptual Evaluation of Own Voice and Social Virtual Reality Using the MPEG-I Immersive Audio Standard. In: Express Paper. Audio Engineering Society; 2025. Paper 376. Available from: https://aes.org/publications/elibrary-page/?id=23050

@inproceedings{Silzle2025_23050,
  author = {Silzle, Andreas and Disch, Sascha and Habets, Emanuel and Rose, Sharon and Dennerlein, Felix and Adami, Alexander and Lieb, Thomas},
  title = {{Perceptual Evaluation of Own Voice and Social Virtual Reality Using the MPEG-I Immersive Audio Standard}},
  note = {Paper 376},
  year = {2025},
  month = oct,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=23050}
}

TY  - CPAPER
TI  - Perceptual Evaluation of Own Voice and Social Virtual Reality Using the MPEG-I Immersive Audio Standard
AU  - Silzle, Andreas
AU  - Disch, Sascha
AU  - Habets, Emanuel
AU  - Rose, Sharon
AU  - Dennerlein, Felix
AU  - Adami, Alexander
AU  - Lieb, Thomas
M1  - Paper 376
PY  - 2025
DA  - 2025/10/14
UR  - https://aes.org/publications/elibrary-page/?id=23050
PB  - Audio Engineering Society
LA  - en
AB  - The recently finalized MPEG-I immersive audio standard provides 6-degrees-of-freedom audio rendering for any kind of virtual-reality acoustic scenarios and applications. Since the standard supports low-latency real-time auralization of incoming speech signals including ones own voice in a virtual acoustic scene, it is well suited for future Social VR applications. This paper describes dedicated technical and perceptual implications that are important to consider in the design of such Social VR applications, it discusses the critical influence of latency of the implementation and provides a quantitative evaluation of the major factors that are relevant for a satisfactory user experience. Investigated factors include timbre, reverberation, signal latency and overall plausibility. A perceptual evaluation of these factors is performed. Therefore, a prototype Social VR simulation of a two-client system that allows communication in a virtual environment with the MPEG-I audio renderer including the visual and acoustic representation of the respective other users by an avatar is used. As an addition to the MPEG-I audio renderer, different acoustic frontend software modules, such as automatic gain control and dynamic range compression for level adjustments, dereverberation and denoising were used to pre-process the incoming voice signals picked up by the headsets microphones. Different configurations of sound level, noise level and reverberation were evaluated. The pre-processed remote voice was generally rated more plausible. In comparison, the test subjects found the non-pre-processed version of their own voice more plausible than the pre-processed one. The reasons for this are discussed.
ER  -