In video conferences, it is important to perceive the other party as if they were physically present there in order to communicate naturally. Our goal is to develop an acoustic system that can reproduce the position and orientation of the remote speaker. Wave field synthesis (WFS) is a well-established technique for achieving this, as it enables spatial audio to be reproduced with great accuracy. However, WFS is often difficult to implement in practice because it requires many microphones and loudspeakers. To address this issue, we focus on the limitations of human auditory perception in order to create a simpler yet effective acoustic system for conversations. In this study, we propose placing speaker arrays, each consisting of two angled loudspeakers, at equal intervals. According to our previous research, this configuration and its control would enable sound radiation in any direction within a 180° range from any position between the arrays. We conducted numerical simulations to confirm this. The results suggested that setting the distance between the arrays at a certain level or below would ensure accurate interaural level differences and, consequently, proper presentation for listeners. This knowledge is useful for achieving a practical balance between system performance and simplicity.
Kishimura et al. (Wed,) studied this question.