Contribution
Perceptually Validated Auralization with the Chalmers Auralization Toolbox
* Presenting author
Abstract:
We present the Chalmers Auralization Toolbox, which is an open-source collection of methods for binaural auralization of simulated sound fields. Such simulation data are typically obtained from wave-based simulation methods like FDTD or BEM. The input data to the toolbox is the sampled sound field, and different arrangements of the sampling points can be handled. We identified in a formal perceptual study that spherical surface sampling grids are the most preferrable arrangement because they allow for achieving perceptually transparent auralization with the lowest number of sampling points, which is 289 nodes of sound pressure and particle velocity. We speak of perceptually transparent auralization when the binaural signals from the auralization method are perceptually indistinguishable from the ground truth binaural signals. We discuss the potential of the toolbox for auralization of acoustic simulation data that originate from methods other than wave-based ones such as geometric acoustics-based or radiocity-based ones.