Data Sonification as a Scientific Tool: How Astronomers Turn Light Curves into Music

Data sonification, the translation of numerical data into sound, has emerged over the past decade as a serious methodology in astronomy, driven by the fields' dependence on time-series data, the accessibility needs of blind and low-vision researchers, and the sheer scale of astrophysical data. This review synthesizes empirical and technical literature on astronomical sonification to evaluate whether it functions as a genuine scientific tool rather than merely an artistic or outreach exercise. It surveys the three dominant techniques and examines controlled signal-detection studies, accessibility and public-engagement surveys, and emerging machine-learning applications. The evidence indicates that sonification performs comparably to visual analysis at high signal-to-noise ratios and narrows the performance gap between expert and non-expert listeners but underperforms at low signal-to-noise ratios and exhibits a consistent liberal response bias toward false positives. Sonification has also demonstrated concrete value in accessibility, education, and public engagement, and in at least one documented case revealed a scientific feature that visual inspection had missed. However, the field remains constrained by the absence of standardized mapping conventions and reporting practices, small and inconsistent sample sizes, and a persistent gap between accessibility claims and genuine co-design with blind and low-vision communities. As data volumes from next-generation surveys such as the Vera Rubin Observatory and the Square Kilometer Array exceed human listening capacity, machine learning and standardization are identified as the critical next steps for sonification to move from a proof-of-concept tool to a routine component of the astronomer's toolkit.