When wireless headphones equipped with a head-tracking system are connected to a PC or smartphone, the head-related impulse responses (HRIRs) used to synthesize 3-D audio should be updated instantaneously in response to the listener’s head movements. However, this update is often delayed due to a latency over wireless connections, resulting in reduced plausibility of the audio scene. To address this issue, we investigated a binaural rendering system that is not affected by transmission delay. In this approach, several dozen audio signals to be rendered are first down mixed into five to eight channels on the transmitter side using panning techniques, and then transmitted to the headphones. Binauralization is performed locally within the headphones using the down mixed signals, based on the directions of the signals relative to the listener’s head orientation. Because the binauralization process occurs independently of the wireless connection, transmission delay does not impact the synthesis of 3-D audio. To prevent degradation in audio quality caused by rotating panned signals during binauralization, panning parameters are calculated for multiple possible head rotation angles on the transmitter side, and the results are averaged. Through this process, localization performance is maintained, even in the presence of head movement during rendering.
Nishiguchi et al. (Wed,) studied this question.