13th Speech in Noise Workshop, 20-21 January 2022, Virtual Conference 13th Speech in Noise Workshop, 20-21 January 2022, Virtual Conference

P25 Investigating auditory selective attention using musical scenes: effect of distraction by concurrent melodies.

Mathilde de Kerangal
Imperial College London, London, UK

Tobias Reichenbach
Friedrich-Alexander-University Erlangen-Nürnberg, Erlangen, Germany

(a) Presenting

Auditory scene analysis is a fundamental aspect of music perception. Polyphonic music consists of the superposition of independent melodic streams and can be used as a tool to study the perception of multi-source scenes. The current study aimed at imitating the challenges of a complex environment (like a cocktail party) and, for that purpose, used musical stimuli in a selective attention task. More specifically, the effect of distraction by concurrent melodies was investigated by (1) increasing the number of distractor melodies and (2) varying the distance in frequency between the distractor and the target melodies. In the first experiment, the participants (N=15) were presented with a selection of chorales of JS Bach containing 1, 2 or 4 parts playing concurrently. The task was to track a target melody - the “leading” part (highest-pitched melody) or the “bass” part (lowest-pitched melody) - and indicate each time two consecutive notes had the same pitch. The performance decreased as the number of distracting melodies increased. Additionally, this decrease was sharper when the subjects were instructed to focus on the bass part; in other words, participants were more distracted when they had to ignore higher pitched melodies compared to lower pitched ones. These effects could be explained by the variation of signal to noise ratio resulting from the addition of auditory sources. They could also be explained by an effect of density related to the addition of sources close to the target stream – a second experiment was designed to isolate this effect of density. In the second experiment, the participants (N=15) were presented with chorales containing 2 parts - the leading and the bass part. The distance in frequency between the two parts was systematically varied resulting in a “dense” and a “sparse” condition. In the dense condition, the distracting part was moved one octave up or down so that it was closer to the target melody; the sparse condition included the leading and bass parts initially present in the chorales. The performance decreased in the dense condition compared to the sparse one, demonstrating an effect of informational masking related to the distance to the target melody. The effect was similar when the participants had to focus on the leading and bass melodies. In summary, the current study shows that the number of sources as well as the frequency distance between sources impacts our ability to selectively attend to a target sound.

Last modified 2022-01-24 16:11:02