From CS 294-10 Visualization Sp10
- Yotam Mann
Traditional music notation gives the familiar viewer a good understanding of the music that it represents in a static way. In the later half of the 20th century, and certainly so far in the 21st century, there has been a trend away from traditional music scores and notation. Many composers, DJs, recording artists, etc. rely entirely on the waveform representation of sound that is provided with most digital audio workstations. What is given in these representation is the amplitude of the sound at any point in time, but little else. I would like to create a representation of recorded sound that would better convey the sound of that material.
An inspiration for this kind of work is the highly stylized graphic scores of Ligeti, especially Artikulation which was created by Rainer Wehinger. This score, in contrast with traditional scores, was created after the composition was complete; it merely serves as a graphic representation of the sound file. At the moment, it is impossible to turn recorded sound back into music notation with a computer, but it is possible to visualize certain elements of the sound that can be extracted with the FFT of the audio file. There is a growing field of Music Information Retrieval which does just that, and some free tools that extract data from audio files. I plan to create a visual language for audio inspired by Wehinger's and incorporate that into the digital audio workstation Ableton Live in place of its traditional waveform representation.
Initial Problem Presentation
Midpoint Design Discussion
- Link to slides here
- File:Final Presentation.zip
- Since the software requires Ableton LIve and Max/MSP, I have not posted it, but included videos of the visualizations in my presentation.