Article
Adapting Sound Reproduction to Listener Position with Dynamic Loudspeaker Equalization (en)
* Presenting author
Abstract:
Virtual acoustic environments and free-field reproduction remove the need for individualized HRTFs and enable experiments with hearing aid or cochlear implant wearers. However, the reproduction accuracy of conventional sound field synthesis techniques decreases with increasing distance to the center of the loudspeaker array, the typical listener position. Kuntz and Seeber (Forum Acusticum 2023) showed the possibility to move the Ambisonics sweet-spot inside the loudspeaker array by changing the point where loudspeaker equalization is computed. At 2 kHz, sweet-spot sizes of at least 22 cm were achieved for listener positions up to 1.4 m from the center. In this study we investigate the use of real-time convolution to dynamically adapt the loudspeaker equalization filters to a tracked listener position. An equalization filter look-up table is built from loudspeaker directivity measurements. Attenuation and time of flight changes are compensated following the 1/r decay. First measurements show that the broadband level can be kept within ±1 dB of the target. Informal listening showed that dynamic artefacts were inaudible for speech, but were still audible for noise at high frequencies due to the currently low update rate.