Visualizing Music in its Entirety using Acoustic Features: Music Flowgram

In this paper, we present an automatic method for visualizing a music audio file from its beginning to end, especially for classical music. Our goal is developing an easy-to-use visualization method that is helpful for listeners and can be used for various kinds of classical music, even for complex orchestral music. To represent musical characteristic, the method uses audio features like volume, onset density, and auditory roughness, which describe loudness, tempo, and dissonance, respectively. These features are visually mapped into static two-dimensional graph, so that users can see how the music changes by time at a look. We have implemented the method with Web Audio API, and it works on recent version of web browsers like Chrome, Firefox, Safari, and Opera. Users can access to the visualization system on their web browser and make visualizations from their own music audio files. Two types of user tests were conducted to verify effects and usefulness of the visualization for classical music listeners. The result shows that it helps listeners to memorize and understand a structure of music, and to easily find a specific part of the music