Perceptually motivated measurement of spatial sound attributes for audio-based information systems

EPSRC Project Reference: GR/R55528/01
Start date: 1 February 2002
End date: 30 April 2005

Principal Investigator: Dr Tim Brookes
Co-Investigator: Dr Francis Rumsey
Research Fellow: Dr Russell Mason
Research Assistant: Dr Tobias Neher

Industrial Partners: BBC Research and Development, Adrian James Acoustics

Background

In consumer entertainment, virtual reality, video conferencing and multimedia computer systems where an audio element is involved, the ultimate judge of sound quality is the human listener. This means that a large proportion of research into sound quality currently involves subjective listening tests, which are expensive and time consuming. In addition, it requires the training of an expert listening panel if consistent and reliable results are to be obtained. As an alternative to this, a perceptually valid objective measurement technique would be more accurately repeatable than subjective tests, and would save time and money [Grewin 1995].

Sound quality is a multidimensional information artefact that incorporates a number of perceived quality attributes. One set of these attributes relate to the spatial impression of the sound, a factor that has been important in concert hall acoustics for a number of years [Ando 1985]. With the growing number of audio systems that can deliver enhanced spatial audio, it is also becoming increasingly important for reproduced and synthesised sound [Toole 1985]. A number of measurements that relate to the spatial properties of acoustical environments have been proposed [Beranek 1996], however, no measurement has been found to be a precise predictor of subjective spatial perception for a wide range of types of sounds.

Project aims

This project aims to investigate objective measurements that relate to certain attributes of the perceived spatial impression of a sound. The project is structured to approach this problem by undertaking subjective experiments that investigate the perceived effect of controlled stimuli in as much detail as is possible. From this, it is expected that a measurement model can be derived that will be applicable to a wide range of sounds and a wide range of acoustic, electronic or electro-acoustic systems, including room acoustics, concert hall acoustics, public address system design, sound recording, sound synthesis, sound reproduction, and video conferencing In addition to this, the research that is undertaken will contribute to our knowledge of the human hearing system, which will enable artificial signals with prescribed properties to be created more accurately, and will help with the synthesis of artificial sounds and acoustical environments in a wide range of areas including virtual reality, multimedia computing and sound recording.