Computer scientists aim to identify link between places and sounds

soundsA study being carried out by the Universities of Aberdeen, Stirling and Edinburgh is aiming to identify and analyse sounds associated with particular places in a bid to gain an insight into the impact of an environment on mood and memory.

People associate places – such as their home or workplace – with particular sounds. However, it has been difficult for academics to investigate the link between the two due to the complexity of sound. Existing methods typically involve listening to long recordings, gathering information subjectively or using frequency analysis.

But computer scientists at Stirling, in partnership with academics from Aberdeen and Edinburgh Universities, have developed a tool that uses feedback from large groups of people to identify sounds commonly associated with particular places.

Four sites across Stirling and Aberdeen – comprising each campus and both city centres – are the focus of the project. The sounds will be analysed to discover how they develop over time, sound collections reflecting each place will be produced and participants – who will hear sounds for the sites via a web page a few times over a three-month period – will have their mood tracked.

The team believes its work on mood and memory could provide a valuable insight into supporting groups of people, such as those living with dementia.

“The key aim for the project was to examine the apparent link between people’s memory of or familiarity with a place and sound through a methodology that has not been widely used in sound studies or sonic arts,” said Dr Suk-Jun Kim, of Aberdeen University and SERG, its sound and electro-acoustic music research group.

“This link is ‘apparent’ mainly because such an examination of the connection between place and sound has often been explored by sound artists or scholars of sound studies who have approached the link in a more artistic, intuitive or conceptual level. In this project, however, we wanted to incorporate a different method – an evolutionary algorithm to sift through a vast amount of sounds with the help of participants.”

Kim and his PhD student Jamie Lawson collected sounds from the city centre and the university both in Aberdeen and Stirling and cut them into 3-second samples “so that there may be just enough sonic information for participants to recognise a place in them”. The sound samples were then integrated into a web app developed at Stirling University.

The team now needs as many participants as possible who are familiar with either of the cities to participate in this project, as the evolutionary algorithm requires a large amount of data for it to work best.

“This is a pilot project, meaning that not only are we interested in the results of this specific question about the link between place and sound, but we are also hoping to find out the feasibility and practicality of the methodology we devised in sound studies,” added Kim. “If successful, we hope to use this methodology for asking bigger issues such as mental health and sound.”

Dr Alexander Brownlee, a computing science research fellow at Stirling, commented: “Sound, memory and emotion are tightly interwoven, and often a person will associate a place with particular sounds. Artists have tried to identify the identity of place through sound, but this is challenging because the sounds are too complex, and elicited sounds are inevitably biased by preconceptions of the artist.

“We are interested in exploring the relationships between sound and places for groups of people, so we’ve developed a tool that will use the feedback from many people to find sounds that are commonly associated with particular places.”

The team has already collected more than 16,000 sounds from the four areas and hope the next stage, involving public participation, will produce interesting results. Volunteers are asked to sign up here , listen to sounds and provide feedback.