Early auditory stream formation of simultaneous musical objects

Poster Session A - Saturday, April 13, 2024, 2:30 – 4:30 pm EDT, Sheraton Hall ABC

Shu Sakamoto¹ (sakams1@mcmaster.ca), Emily Wood¹, Laurel J. Trainor¹; ¹McMaster Institute of Music and the Mind, McMaster University

Our brains evolved to parse multiple simultaneously-sounding objects into representative streams, based on spectral, temporal, and spatial information. In the case of multiple talkers, we form a stream for each person, and these streams cannot typically be combined if sematic meaning is to be retained. In music, on the other hand, different streams (e.g., instruments) can be combined meaningfully. Here we investigated the role of attention on early integration of streams in music. In particular, we trained a model to reconstruct the stimulus from the measured EEG response to test whether initial early stream formation occurs automatically regardless of attention aimed at stream integration. Participants (n=22) were presented with two sources of naturalistic piano music (presented 30 degrees to the right and left) that could be perceived as two separate pieces or combined into a meaningful whole. They were instructed to attend to one and ignore the other (segregate the pieces) or to attend to both (integrate). We analyzed how well their EEG tracked each of the two pieces independently as well as the combined mixture of sound. We found that the brain tracked the spectro-temporal modulation of each piece independently in the early processing stage (lag of 50–80 ms between sound and the corresponding EEG response), not only when they aimed their attention to segregate, but also to integrate the two pieces. This suggests that stream formation for music occurs automatically at an early stage of processing, regardless of overt attention to actively segregate or integrate the streams.

Topic Area: ATTENTION: Auditory