Effect of Interpolation Artifacts on Perceived Stability of Nearby Sources in a Navigable Reverberant Virtual Environment

De Muynke J, Poirier-Quinot D, Katz BFG. Effect of Interpolation Artifacts on Perceived Stability of Nearby Sources in a Navigable Reverberant Virtual Environment. J Audio Eng Soc. 2024;72(10):664-678. doi:10.17743/jaes.2022.0158

@article{DeMuynke2024_22770,
  author = {De Muynke, Julien and Poirier-Quinot, David and Katz, Brian F. G.},
  title = {{Effect of Interpolation Artifacts on Perceived Stability of Nearby Sources in a Navigable Reverberant Virtual Environment}},
  journal = {Journal of the Audio Engineering Society},
  volume = {72},
  number = {10},
  pages = {664--678},
  year = {2024},
  month = oct,
  publisher = {Audio Engineering Society},
  doi = {10.17743/jaes.2022.0158},
  url = {https://doi.org/10.17743/jaes.2022.0158}
}

TY  - JOUR
TI  - Effect of Interpolation Artifacts on Perceived Stability of Nearby Sources in a Navigable Reverberant Virtual Environment
AU  - De Muynke, Julien
AU  - Poirier-Quinot, David
AU  - Katz, Brian F. G.
T2  - Journal of the Audio Engineering Society
J2  - J. Audio Eng. Soc.
VL  - 72
IS  - 10
SP  - 664
EP  - 678
PY  - 2024
DA  - 2024/10/15
DO  - 10.17743/jaes.2022.0158
UR  - https://doi.org/10.17743/jaes.2022.0158
PB  - Audio Engineering Society
LA  - en
AB  - Convolution with spatial Room Impulse Responses can achieve realistic auralizations. When combined with interpolation between spatially distributed RIRs, this technique can be used to create navigable virtual environments. This study explores the impact of various interpolation parameters on the perceived auditory stability of a nearby static sound source with listener movements in a reverberant environment. The auditory scene was rendered via third-order Ambisonic RIR convolution combined with magnitude-least-squares binaural decoding using nonindividualized head-related transfer functions. First, the estimated direction of arrival as a function of the listener’s position within a 2D grid of RIRs under various configurations is examined as an objective metric. The perceived stability of the auditory source is then assessed through a perceptual experiment. Participants freely explored a virtual scene reproduced over headphones and a tracked head-mounted display. They were asked to rate the stability of a nonvisual source under various conditions of RIR grid density, interpolation panning method, and reverberation time. Results indicate no need to use an RIR grid size finer than 1m to optimize source stability when using a three-nearest-neighbor interpolation scheme.
ER  -