The Potential Role of Listening Modes in Auditory Interfaces for Location-based Services

The auditory modality offers several advantages as a means of communication for the purposes of location-based services (LBS), including fast response time [1], low processing and storage overheads [2], and hands/eyes-free mobility. However, with more and more sound-producing technology being used in day-to-day life, the battle for our acoustic attention has led to a steady rise in acoustic noise levels [3]. In an already noisy environment, it is tempting for the sound designer to simply use more volume as a means of gaining the listener’s attention but this only serves to create a vicious circle of noise in which every sound designer is merely struggling to be heard over the noise of every other sound designer. The field of soundscape theory, however, may offer some potential solutions in this regard. Soundscape theory, as described by Schafer [4, 5] and Truax [6], considers sound from a more holistic point of view and the concept of listening modes considers the different levels of attention we pay to auditory stimuli depending on context and location within the soundscape. While several different theoretical listening modes have been proposed across the various acoustic disciplines, there is a need for empirical data to support the existence of these modes. One area in which there is a certain amount of empirical data is in relation to spectral bandwidth and what Krause [7-9] has called his ‘niche theory’. Niche theory describes the way in which different species appear to occupy discrete frequency bandwidths within the soundscapes of natural habitats; it is argued that this natural balance keeps redundant noise to a minimum and enables more efficient acoustic communication. If the principles observed in niche theory were to be observed in human listening behaviour, a new approach to sound design might be possible whereby auditory stimuli exploit specific frequency bandwidths in order to maximise information exchange without necessarily raising noise levels. In this paper, we outline a proposed experiment in which listeners are asked to engage in a foreground task that encourages competitive conversation while also attending to a background listening task in which participants have to acknowledge background non-speech sounds of varying spectral bandwidth presented at random intervals. Our aim is to compare the spectrogram information of both the foreground task and the background stimuli to see if relative spectral bandwidth has any discernible effect on stimulus identification success rate and response time.