Efficient Rendering Of HOA Using HOS And Magnitude Least Squares

Hollebon J, Fazi FM. Efficient Rendering Of HOA Using HOS And Magnitude Least Squares. In: AES 2024 International Conference on Audio for Virtual and Augmented Reality. Audio Engineering Society; 2024. Paper 28. Available from: https://aes.org/publications/elibrary-page/?id=22677

@inproceedings{Hollebon2024_22677,
  author = {Hollebon, Jacob and Fazi, Filippo Maria},
  title = {{Efficient Rendering Of HOA Using HOS And Magnitude Least Squares}},
  booktitle = {AES 2024 International Conference on Audio for Virtual and Augmented Reality},
  note = {Paper 28},
  year = {2024},
  month = aug,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=22677}
}

TY  - CPAPER
TI  - Efficient Rendering Of HOA Using HOS And Magnitude Least Squares
AU  - Hollebon, Jacob
AU  - Fazi, Filippo Maria
T2  - AES 2024 International Conference on Audio for Virtual and Augmented Reality
M1  - Paper 28
PY  - 2024
DA  - 2024/08/05
UR  - https://aes.org/publications/elibrary-page/?id=22677
PB  - Audio Engineering Society
LA  - en
AB  - Higher-Order Stereophony (HOS) expands traditional sterophonic techniques to higher order soundfield reproduc-tion, in a similar manner to Higher-Order Ambisonics (HOA). HOS reproduces the soundfield accurately along a line, across the listeners interaural axis only. This is equivalent to reproducing a soundfield using just the m = 0 spherical harmonics. HOS only requires (N +1) channels for order N reproduction, compared to (N +1)2 for HOA, a significant reduction particularly when applied to binaural rendering where each channel requires a convolution process. This work presents a signal flow to decode HOA signals to an equivalent HOS representation utilising the spherical harmonic representation, then renders binaural output using HOS rendering. Head-Related Transfer Functions (HRTFs) rendered using the decoder are compared to the full reference HOA case. The rigid sphere HRTF is shown to be fully reproduced for all directions using the HOS renderer. Next, utilising the established link to Ambisonics, the high frequency mitigation approach magnitude least squares is also adapted to work with the HOS framework to reduce high frequency errors when the HRTFs are sampled to a higher order than the input soundfield.
ER  -