Individualized Binaural Technology

Dr. Bruno Masiero (RWTH Aachen University, Germany)

APPLIED SIGNAL PROCESSING SERIES

DATE: 2013-05-17
TIME: 11:00:00 - 12:00:00
LOCATION: CSIT Seminar Room, N101
CONTACT: JavaScript must be enabled to display this email address.

ABSTRACT:
In this work the importance of individualization in binaural technique is investigated. The results extend the present knowledge on the efficient measurement of individual head-related transfer functions (HRTFs) and highlight the importance of individual equalization filters in binaural re-production, both via loudspeakers and headphones. Moreover, an integrated framework for the calculation of such equalization filters is presented.

An innovative measurement setup was developed to allow the fast acquisition of individual HRTFs. The hardware was designed to be compatible with the range extrapolation technique, which makes the description of the HRTF's distance-dependence possible. Major speedup was obtained by optimizing the multiple exponential sweep method. An individual HRTF dataset with approximately 4000 directions can be measured in less than 6 minutes with this new setup.

Crosstalk cancellation (CTC) filters are required when playing back binaural signals via loud-speakers. To allow listeners to freely move their heads, switching between multiple loudspeakers is required and the CTC filters must be continuously updated according to the tracked head position. Filter calculations are carried out in the frequency domain for speed reasons. To impose causality constraints to the regularized frequency domain calculations, a CTC filter calculation framework is proposed, which incorporates a new approach for the multi-channel minimum-phase regularization. This framework also addresses the switching between active loudspeakers through the use of a weighted filter calculation. A sound localization test showed that individualized CTC systems provide performance similar to that of binaural listening while nonindividualized CTC systems provide a significantly lower localization performance.
BIO:
Bruno Masiero received his B.S. and M.Eng. in electrical engineering from the University of SAo Paulo, Brazil, in 2005 and 2007 respectively,and his doctoral degree in acoustic engineering from the RWTH Aachen University, Germany,in 2012. His research interests include acoustic signal processing,spatial sound reproduction, and psychoacoustics.