S^3MASH: Spatial Sound Scene Matching using Single-Channel Audio

Gonzalez R, McCormack L, Politis A. S^3MASH: Spatial Sound Scene Matching using Single-Channel Audio. In: AES 2024 International Conference on Audio for Virtual and Augmented Reality. Audio Engineering Society; 2024. Paper 15. Available from: https://aes.org/publications/elibrary-page/?id=22664

@inproceedings{Gonzalez2024_22664,
  author = {Gonzalez, Raimundo and McCormack, Leo and Politis, Archontis},
  title = {{S\textasciicircum{}3MASH: Spatial Sound Scene Matching using Single-Channel Audio}},
  booktitle = {AES 2024 International Conference on Audio for Virtual and Augmented Reality},
  note = {Paper 15},
  year = {2024},
  month = aug,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=22664}
}

TY  - CPAPER
TI  - S^3MASH: Spatial Sound Scene Matching using Single-Channel Audio
AU  - Gonzalez, Raimundo
AU  - McCormack, Leo
AU  - Politis, Archontis
T2  - AES 2024 International Conference on Audio for Virtual and Augmented Reality
M1  - Paper 15
PY  - 2024
DA  - 2024/08/05
UR  - https://aes.org/publications/elibrary-page/?id=22664
PB  - Audio Engineering Society
LA  - en
AB  - This paper describes a novel approach for recording and binaurally reproducing spatial sound scenes using the audio from a single microphone. This is realised by recording the sound scene using both a microphone array, which potentially comprises more affordable and lower quality capsules, and a monophonic microphone, possibly featuring a higher quality capsule. By adopting a perceptually motivated sound-field model and estimating the models spatial parameters, it is possible to define target time-frequency-dependent binaural spatial covariance matrices (SCMs). The actual binaural signals can then be synthesised using an adaptive SCM matching renderer, which takes only the higher-quality monophonic audio signal as input. A perceptual study was conducted to compare this novel processing approach, using a tetrahedral array and an omnidirectional microphone, against binaural renderings achieved through traditional Ambisonic means, when using four- and 32-channel arrays. The results show that, despite utilising only a monophonic signal for the spatialisation, the proposed approach yielded binaural renderings that are perceptually in-between the two conventional Ambisonic array renderings, with regards to their perceived spatial accuracy.
ER  -