Subjective perceptual organization of a complex auditory scene

Abstract

Empirical research on the sequential decomposition of an auditory scene primarily relies on interleaved sound mixtures of only two tone sequences (e.g., ABAB…). This oversimplifies the sound decomposition problem by limiting the number of putative perceptual organizations. The current study used a sound mixture composed of three different tones (ABCABC…) that could be perceptually organized in many different ways. Participants listened to these sequences and reported their subjective perception by continuously choosing one out of 12 visually presented perceptual organization alternatives. Different levels of frequency and spatial separation were implemented to check whether participants' perceptual reports would be systematic and plausible. As hypothesized, while perception switched back and forth in each condition between various perceptual alternatives (multistability), spatial as well as frequency separation generally raised the proportion of segregated and reduced the proportion of integrated alternatives. During segregated percepts, in contrast to the hypothesis, many participants had a tendency to perceive two streams in the foreground, rather than reporting alternatives with a clear foreground-background differentiation. Finally, participants perceived the organization with intermediate feature values (e.g., middle tones of the pattern) segregated in the foreground slightly less often than similar alternatives with outer feature values (e.g., higher tones).

Interaction of spatial source separation, fundamental frequency, and vowel pairing in a sequential informational masking paradigm in Mongolian gerbils.

“…Tolnai et al (2015) showed that binaural cues such as ITDs and ILDs effectively provide for a release from IM in humans using the same experimental paradigm as presented here. Along with frequency differences of pure tone stimuli, also the spatial separation of sounds was shown to enhance the segregation of tone streams (Itatani & Klump, 2017b; Thomassen & Bendixen, 2017). For more complex and speech-like stimuli the spatial separation of sounds was also observed being advantageous: spatial cues were observed to induce stream segregation for speech-shaped noise (David et al, 2015) and the spatial separation of sounds enhances speech intelligibility (e.g., Peissig & Kollmeier, 1997).…”

“…Vowels can be segregated by differences in fundamental frequency (Carlyon & Gockel, 2008; David, Lavandier, Grimault, & Oxenham, 2017; Grimault, Micheyl, Carlyon, Arthaud, & Collet, 2000; Hawley, Litovsky, & Culling, 2004; Shackleton & Meddis, 1992; Shackleton, Meddis, & Hewitt, 1994; Vliegen & Oxenham, 1999), but also vowels sharing the same fundamental frequency can be segregated based on their spectral shape (Assmann & Summerfield, 1989, 1990; Culling & Darwin, 1993; Micheyl & Oxenham, 2010). For sequential grouping, the spatial separation of sounds also is important (David, Lavandier, & Grimault, 2015; Hartmann & Johnson, 1991; Middlebrooks & Onsan, 2012; Moore & Gockel, 2012; Peissig & Kollmeier, 1997; Saupe, Koelsch, & Rübsamen, 2010; Thomassen & Bendixen, 2017). Speech intelligibility improves when target speech streams are spatially separated from competing speech (Hawley, Litovsky, & Colburn, 1999; Hawley et al, 2004), and the importance of spatial separation of concurrent sounds increases with the increasing complexity of an auditory scene (Kidd, Arbogast, Mason, & Gallun, 2005; Yost, Dye, & Sheft, 1996).…”