Selecting birdsongs for auditory masking: a clustering approach based on psychoacoustic parameters

• Main Authors: Ong, Zhen-Ting, Lam, Bhan, Hong, Joo Young, Ooi, Kenneth, Gan, Woon-Seng
• Other Authors: School of Electrical and Electronic Engineering
• Format: Conference or Workshop Item
• Language: English
• Published: 2023
• Subjects: Science::Physics::Acoustics; Social sciences::Psychology::Affection and emotion; Soundscape; Masking; Birdsongs; Clustering; Psychoacoustic Parameters
• Online Access: https://hdl.handle.net/10356/170483
• Institution: Nanyang Technological University

Birdsongs are widely reported as effective auditory maskers to enhance soundscape quality and even reduce the perceived loudness of unwanted sounds. However, the bird species are usually under-reported and chosen arbitrarily. The ambiguity about the objective characteristics of the birdsongs casts doubt on the generalisations of those studies. To narrow down the selection of birdsongs for a subjective study, we propose a method to cluster birdsongs based on psychoacoustic parameters. In total, birdsongs from 28 bird species (10 s), set to the same level, are used in this study. The samples are analysed in terms of psychoacoustic parameters such as, loudness, sharpness, roughness and fluctuation strength. Based on the calculated psychoacoustic parameters, principal component analysis (PCA) and hierarchical cluster analysis (HCA) for the birdsongs are conducted. The results of HCA show that the birdsongs are classified into five clusters based on the psychoacoustic parameters. In addition, PCA results revealed that the temporal variance of sharpness and loudness are the critical factors to discriminate the five clusters of birdsongs.