Meeting Topic: Binaural Sound Reproduction over Loudspeakers
Moderator Name: Piotr Majdak
Speaker Name: Bruno S. Masiero
Meeting Location: Seminar room of the Institute for Acoustic Research, Austrian Academy of Sciences, Vienna
It is long known that to correctly reproduce a binaural signal through a pair of loudspeakers, this signal has to be pre-filtered to compensate for the crosstalk effect that will otherwise ruin the spatial clues contained on the binaural signal. In this work we make a brief review of the design technique for crosstalk cancellation (CTC) filters specially focused on its application in virtual reality installations and discuss still open dilemmas.
CTC filters are designed based on the transfer paths between loudspeakers and listener ears and usually delivers channel separations of over 20 dB in a small region around the listeners head, the so-called sweet spot. If the listener moves away from the sweet spot, channel separation will deteriorate and the spatial cues from the binaural signal will be lost. A great deal of researchers tried to answer the question on how to increase the size of the sweep spot through a more skilled system setup, specially important for virtual reality applications where the users should be allowed to freely move and rotate their heads.
For such applications dynamic CTC systems can be implemented with the aid of a head-tracking device to determine distance and orientation of the listener's head to the loudspeakers, allowing a constant update of the cancellation filters.
An often debated question is whether individual head-related transfer functions (HRTF) are required as plant model for the CTC filter generation or if a simplified plant model will deliver sufficient channel separation for a wide range of listeners.