Updates on the Real-Time Spherical Array Renderer (ReTiSAR)

Paper in proceeding, 2020

We recently presented ReTiSAR, a framework for binaural rendering of spherical microphone array signals in real-time. The array signals and the employed head-related transfer functions are convolved in the spherical harmonics domain to compute the resulting ear signals and virtually place a listener into the captured sound field. In this contribution, we present the latest additions to the Python software package. These comprise, among others, an interface to the Spatially Oriented Format for Acoustics (SOFA), the ability of switching the spherical harmonics rendering order during runtime as well as performance optimizations enabling real-time rendering up to 12th-order. Furthermore, we integrated enhancements recently proposed in the literature for the upper frequency range where spatial undersampling occurs.