Comparison of 3D Audio Reproduction Methods Using Hearing Devices

Simon LSR, Dillier N, Wüthrich H. Comparison of 3D Audio Reproduction Methods Using Hearing Devices. J Audio Eng Soc. 2020;68(12):899-909. doi:10.17743/jaes.2020.0059

@article{Simon2020_21005,
  author = {Simon, Laurent S. R. and Dillier, Norbert and Wüthrich, Hannes},
  title = {{Comparison of 3D Audio Reproduction Methods Using Hearing Devices}},
  journal = {Journal of the Audio Engineering Society},
  volume = {68},
  number = {12},
  pages = {899--909},
  year = {2020},
  month = dec,
  publisher = {Audio Engineering Society},
  doi = {10.17743/jaes.2020.0059},
  url = {https://doi.org/10.17743/jaes.2020.0059}
}

TY  - JOUR
TI  - Comparison of 3D Audio Reproduction Methods Using Hearing Devices
AU  - Simon, Laurent S. R.
AU  - Dillier, Norbert
AU  - Wüthrich, Hannes
T2  - Journal of the Audio Engineering Society
J2  - J. Audio Eng. Soc.
VL  - 68
IS  - 12
SP  - 899
EP  - 909
PY  - 2020
DA  - 2020/12/06
DO  - 10.17743/jaes.2020.0059
UR  - https://doi.org/10.17743/jaes.2020.0059
PB  - Audio Engineering Society
LA  - en
AB  - In order to study 3D auditory perception for hearing aid users, spatial audio reproduction systems have been considered. Numerous solutions were designed in the past, but few studies evaluated the effect of 3D audio reproduction artifacts on hearing devices. This study focuses on loudspeaker-based 3D audio reproduction and compares the effects of distance-based amplitude panning (DBAP), third and fifth order Ambisonics (HOA), multiple-direction amplitude panning (MDAP), and vector-base amplitude panning (VBAP) on the 3D directivity index (DI) of hearing device beamformers, on interaural time differences (ITD), and on interaural level differences (ILD). Measurements were conducted at the sweet spot and for two off-center positions in a non-anechoic room using 32 loudspeakers. Results show that the effect of the reproduction system on the DI at the sweet spot was larger for third order Ambisonics and MDAP and smaller for fifth order Ambisonics. They also show that DBAP cannot reproduce ITDs correctly and that the ILDs obtained with HOA reproduction contain larger errors than with the other reproduction systems. Finally the results show that the DI, ITD, and ILD errors increase with the distance to the sweet spot and that this increase is larger for HOA than for the other reproduction methods.
ER  -