Your Location Within Site:
Surround sound
Surround sound systems aim to produce a desired 3D sound field over a region of space using an array of loudspeakers. They require a large number of loudspeakers for reproduction over a large region at full audio bandwidths. Inevitable compromises must be made for practical installations, and this leads to the need to optimise the system to minimise the reproduction errors that occur with a limited number of loudspeakers.
The theory of 3D surround sound is based on the solution of the wave equation in spherical coordinates, and involves the use of spherical harmonics, an approach known as Higher Order Ambisonics.
Zeroth and first order spherical harmonics are the monopole and three dipoles
Second order spherical harmonics produce finer variation with angle.
An alternative general approach to sound reproduction based on the application of the Kirchhoff Helmholtz integral equation is known as Wave Field Synthesis. When used in surround systems the alternative approach can be shown to be equivalent to Higher Order Ambisonics, but there are differences in the practical application.
Lower cost surround sound systems use a 2D array of loudspeakers to produce a 2D sound field. A simple example is the 5.1 surround system used in home theatres. These systems present significant challenges to high quality listening due to the irregular angular spacing of the loudspeakers, and the small region over which accurate reproduction is possible.
3D sound fields may be recorded by using a microphone array which carries out a spherical harmonic decomposition of the sound field, producing a set of audio (ambisonics) signals that describe the 3D field. One of the practical aspects of recording arrays is the effects of phase and amplitude mismatch in microphones, which distort the desired spherical harmonic responses and produce limitations on accuracy for a given size of array and number of transducers. One of our research goals is the development of recording arrays that can correct for mismatch errors.
The required loudspeaker signals for surround reproduction are obtained in general from a filtered sum of the recorded ambisonics signals. The precise combination of these signals depends on the loudspeaker array geometry. For plane wave reproduction, the loudspeaker signals are simply obtained as weighted sums of the recorded ambisonics signals.
Surround sound systems are designed to operate in rooms which have absorbing walls, which prevent sound generated by the loudspeakers from reflecting from the room surfaces. In practise, however, surround systems operate in rooms with non-absorbent walls. The resulting reflections produce an unwanted reverberant sound field inside the surround sound array which corrupts the desired sound field. Surround systems can compensate for room acoustics by measuring the spherical harmonic responses produced by each loudspeaker at the center of the array and implementing a pre-processor that allows cancellation of room effects. However these systems are complex and sensitive to errors in the measurement. We are examining methods for reducing room effects using more robust signal processing, and directional loudspeakers.