Analysis of Sonic Effects of Music from a Comprehensive Datasets on Audio Features

Abstract

Music, for the longest time, has impacted human lives tremendously. The ability of music to access and activate a wide range of human emotions is sensational. Toward this end, audio features provide a variety of information necessary for sound engineers, music producers, and artists to improve their craft to excite the vast majority of music listeners across the globe. In this paper, analysis of audio features derived using the Spotify web API endpoint and Spotify (Python module for Spotify web servers) is presented. The dataset was curated from audio features of over 160,000 songs released from the year 1921-2020. For clarity, statistical descriptions and probability distribution functions of the audio features are reported. Also, the interrelationship and correlation amongst the various audio features are demonstrated. Overall, the dataset would find useful applications in classical and future music production.