Interpolated HRIRs for System Identification

We simulated time-varying head-related impulse responses (HRIRs) during continuous speaker array rotation using the FABIAN HRIR database, which provides high-resolution measurements for azimuth angles ranging from 0° to 360° and elevation angles from -90° to 90° in an anechoic chamber at a 44.1 kHz sampling rate. HRIRs at 0° head-above-torso orientation (HATO) were interpolated to 0.001° azimuth resolution using inverse distance weighting to simulate a 45°/s rotation. Interpolated HRIRs were included for 12 speakers spaced at 30° elevation intervals (–90° to 90°). A 180° rotation (covering 360° azimuth with the 12 speakers) was simulated to support deep learning-based HRIR identification and head-related transfer function (HRTF) individualization.