Selective listening in music: From psychoacoustical principles to hearing device evaluation

Music typically consists of a flurry of sounds from multiple instruments or voices that overlap in time and frequency. To infer musical structure, listeners, thus, need to select and track sounds from mixtures—a process also known as auditory scene analysis (ASA). ASA is determined by stimulus- and listener-specific properties: While certain instruments or voices naturally stand out in musical mixtures, individual listeners vary in their ability levels to hear out specific elements from mixtures. In the recent work, our team explored the psychoacoustical underpinnings of selective listening in music. In experiments with realistic musical material from popular music, this work identified acoustical features at the basis of vocal salience and specifically highlighted the role of pitch modulation and spectral distinctiveness. Selective listening in orchestral music turned out to be highly dependent on musical context and target sound but less affected by spatial acoustical variables. Using an efficient adaptive test approach, we observed individual differences in selective listening as a function of musical training, age, and hearing loss. The approach was further used to evaluate the performance of hearing devices, providing insight into their effectiveness in enhancing selective listening. Overall, our results underscore the need to develop tailored auditory solutions for listeners with diverse hearing profiles.