Corbyn Langley

Music Visualization Using Graph Theory as a Medium

My project explored various methods for extracting information from music and using this information to trigger visual ques to generate aesthetic images. The end result made use of the FFT and FFT based features to represent music data. A set of rules were created to mandate the iterative creation of graphs. Each iteration represents one second of a song. Each iteration will draw nodes and connect nodes by edges that are colored based on pitch. Similarity features extracted from music are used to connect similar regions within the graphs. This adds additional meaningful structure to the graphs. The resulting graphs convey the structure and dynamical shifts within songs.