Can You Hear Out the Melody? Testing Musical Scene Perception in Young Normal-Hearing and Older Hearing-Impaired Listeners

Abstract

It is well known that hearing loss compromises auditory scene analysis abilities, as is usually manifested in difficulties of understanding speech in noise. Remarkably little is known about auditory scene analysis of hearing-impaired (HI) listeners when it comes to musical sounds. Specifically, it is unclear to which extent HI listeners are able to hear out a melody or an instrument from a musical mixture. Here, we tested a group of younger normal-hearing (yNH) and older HI (oHI) listeners with moderate hearing loss in their ability to match short melodies and instruments presented as part of mixtures. Four-tone sequences were used in conjunction with a simple musical accompaniment that acted as a masker (cello/piano dyads or spectrally matched noise). In each trial, a signal-masker mixture was presented, followed by two different versions of the signal alone. Listeners indicated which signal version was part of the mixture. Signal versions differed either in terms of the sequential order of the pitch sequence or in terms of timbre (flute vs. trumpet). Signal-to-masker thresholds were measured by varying the signal presentation level in an adaptive two-down/one-up procedure. We observed that thresholds of oHI listeners were elevated by on average 10 dB compared with that of yNH listeners. In contrast to yNH listeners, oHI listeners did not show evidence of listening in dips of the masker. Musical training of participants was associated with a lowering of thresholds. These results may indicate detrimental effects of hearing loss on central aspects of musical scene perception.

Listening in the Mix: Lead Vocals Robustly Attract Auditory Attention in Popular Music

“…It is to be noted, that our analysis revealed no systematic differences between participants who declared themselves as musician and those who did not. This held true across all three experiments, even though in previous studies, musicians showed improved results in ASA tasks (e.g., Başkent et al, 2018 ; Madsen et al, 2019 ; Siedenburg et al, 2020 ). The most likely reason to explain this may be that we did not specifically control for an equal number of musicians and non-musicians in a large sample; thus, the proportion of participants considered musicians were only a fraction of the total participants, and therefore the sample size may be too small for an adequate statistical comparison.…”

“…Although no clear effect for a modification of timbral dissimilarity could be observed, the results implied a trend that a reduction of timbral dissimilarity and thus a reduction of acoustical cues lead to a deterioration of stream segregation, further suggesting that a minimum of exogenous cues is necessary to track and separate single streams. In Siedenburg et al (2020) , listeners had to hear out instruments and melodies of varying sound level masked by a simultaneously playing instrument. It was found that participants were able to exploit dips in the masker signal, allowing them to hear the target instrument at lower levels than with a masker that did not contain these dips.…”

Tracking Musical Voices in Bach's The Art of the Fugue: Timbral Heterogeneity Differentially Affects Younger Normal-Hearing Listeners and Older Hearing-Aid Users

“…In the statistical analysis, adding the musical training index to the model did not improve the model fit. Note that previous studies of scene analysis found that musical training improved performance of participants in several scene analysis tasks (see e.g., Madsen et al, 2019;Siedenburg et al, 2020). In the present study, we did not undertake a dedicated comparison of groups of participants with and without musical training, and our musical training index mainly encoded variations in the degree of musical training.…”

“…This project has received funding from the European Union's Framework Programme for Research and Innovation Horizon 2020-2020 under the Marie Skłodowska-Curie Grant Agreement No. 747124.…”

“…More generally, aging has been shown to be associated with a reduced ability in exploiting spectral fine structure cues (Moore, 2015). Recent research on the specific aspect of musical scene analysis has shown that compared to younger normal-hearing listeners, older listeners with moderate hearing impairment have drastically higher thresholds for hearing out a musical melody or instrument in a simple chordal accompaniment (Siedenburg et al, 2020). Whether this reduced ability in scene parsing generalizes toward ecological musical scenarios (i.e., real-world music) remains an open question.…”