Back to list

Transaural Beamforming Methods for Controllable Focused Sound Reproduction

Authors Flock, M.
Year 2009
Thesis Type Diploma thesis
Topic Spatial Audio
Keywords audio reproduction
Abstract In literature, several proposals can be found on how to synthesize or resynthesize sound fields. In general, there are two main strategies. Global approaches, like Wave Field Synthesis or Ambisonics, have been derived from physical equivalents and aim to reproduce a sound field within a whole area. Local approaches on the other hand, are designed to reproduce the sound field at the position of a user only, as for example with binaural signals via headphones. The inherent goal of this work is to elaborate possible improvements for air traffic control communication conditions. In order to free air traffic controllers from the demanding usage of headphones, a loudspeaker-array based application is introduced. The array produces binaural signals at the ears of the user to provide spatialized audio. It therefore is a local approach of sound-source reproduction, that, however, suffers from unwanted cross talk from one binaural channel to the contralateral ear. Different beamforming methods are investigated under the aspect of focusing quality, room excitation, dynamic adaptation and their contribution to an effective transaural rendering. It is a weighted Delay & Sum Beamformer, a Least Squares Beamformer, a Maximum Energy Difference Beamformer and a Minimum Variance Distortionless Response Beamformer. The first method proved to be very feasible for its facile implementation. Although its principle is very simple, its results are comparable to the other methods that use optimization algorithms for the sound-field manipulation. Finally, a cross talk cancellation solution for a loudspeaker array and a binaural signal is deduced and the functionality of the overall system is undermined with measurements results.
Supervisors Sontacchi, A.