Acoustic Structure of Birdsong and Human Mental Wellbeing

BackgroundExposure to birdsong benefits mental health. Different types of birdsongs vary in their capacity to promote wellbeing; however, the underlying acoustic mechanisms are not yet understood.MethodsFrom March 2020 to May 2025, 2,088 participants contributed 11,631 assessments via the smartphone-based Urban Mind platform. Each assessment included a momentary mental wellbeing score and an audio recording of the surrounding environment. Seventy-six acoustic features characterising birdsong structure were extracted and screened via univariable and correlation analyses. Associations between birdsong exposure, acoustic features and momentary mental wellbeing were examined using generalised mixed-effect regression models, adjusting for individual- and context-level covariates.FindingsExposure to birdsong was associated with higher momentary mental wellbeing, with nearly doubled odds of reporting high wellbeing compared with non-birdsong sound environments (adjusted OR = 1·97, 95% CI 1·56–2·50). This association remained robust after adjustment for confounders and showed no meaningful moderation by mental health status, gender, or age. Acoustic analyses identified dynamic spectral variability and temporal intermittency as key features associated with wellbeing. These features were combined into a composite Dynamic Timbre Index (DTI), with higher DTI values consistently associated with better momentary mental wellbeing. Species-level analyses across the Americas, Asia, European Union, and the UK demonstrated convergent acoustic characteristics among high-DTI species.InterpretationThe acoustic structure of birdsong, rather than bird presence alone, is associated with momentary mental wellbeing. Identification of specific acoustic features provides mechanistic insight into how natural soundscapes influence mental health and may inform the development of targeted nature-based interventions. An interactive web-based platform, BirdsongandWellbeing, enables users to compute DTI scores from uploaded audio recordings.