The interactions of spatial and pitch cues in auditory scene analysis

Poster Session E - Monday, April 15, 2024, 2:30 – 4:30 pm EDT, Sheraton Hall ABC

Nima Zargarnezhad¹ (nzargan@uwo.ca), Ezgi Coskun¹, Ingrid Johnsrude¹; ¹Western University

The human auditory system can decompose complex sound mixtures into distinct perceptual auditory objects through a process known as Auditory Scene Analysis (ASA). Pitch and spatial cues are among sound properties known to influence sequential grouping (Plack, 2018). The impact of pitch in ASA has been well characterized using Van Noorden's (1975) ABA paradigm, in which two streams of tones, ‘A’ and ‘B’ are interleaved, and the factors contributing to perceptual segregation of the two streams (including pitch) are explored. Here, we manipulate pitch and spatially displace the A and B sequence sources to understand how spatial and pitch cues trade off in perceptual organization. We used a 91-channel geodesic dome (radius 1.65m) loudspeaker array with 9th-order ambisonic rendering (Sonible, Austria) in a sound-attenuating room. Participants sat in complete darkness with heads at the center, ears at 0 elevation, and eyes fixated on an LED at 0 azimuth. Trials featured 32 repetitions of ABA triplets (A: 125ms, 400 Hz; B:125ms, >= 400 Hz; 125ms inter-triplet interval; 16-sec overall). A and B streams were presented symmetrically on the horizontal plane, and listeners reported segregation/integration via keypresses (Cusack et al., 2004). We first used adaptive testing to create, for each listener, a two-dimensional (space and pitch), 49-coordinate psychophysical space. Then, each coordinate was tested 10 times, measuring the proportion of time streams were segregated. Our results suggest that small spatial separations promote segregation but apparent weightings of the two dimensions in determining segregation vary substantially between individuals.

Topic Area: PERCEPTION & ACTION: Audition