Featur.Ux: Exploiting Multitrack Information For Artistic Visualization
Ireti Olowe, Mathieu Barthet, Mick Grierson, Nick Bryan–Kinns · Goldsmiths (University of London) · 2016
FEATUR.UX (Feature - ous) is an audio visualization tool, currently in the process of development, which proposes to introduce a new approach to sound visualization using pre-mixed, independent multitracks and audio feature extraction. Sound visualization is usually performed using a final mix, mono or stereo track of audio. Audio feature extraction is commonly used in the field of music information retrieval to create search and recommendation systems for large music databases rather than generating live visualizations. Visualizing multitrack audio circumvents problems related to the source separation of mixed audio signals and presents an opportunity to examine interdependent relationships within and between separate streams of music. This novel approach to sound visualization aims to provide an enhanced accession to the listening experience corresponding to this use case that employs non-tonal, non-notated forms of electronic music. Findings from prior research studies focused on live performance and preliminary quantitative results from a user survey have provided the basis from which to develop a prototype that will be used throughout an iterative design study to examine the impact of using multitrack audio and audio feature extraction on sound visualization practice.