Dance Your PhD: Human-Based Percussion and Self-Similarity Detection in Electroacoustic Music

This is a video I edited together and shot along with Andrew Stalick and Nico Galindo for a friend of ours, Anderson Mills, who was the mastermind behind this video submission for the Dance Your PhD contest (gonzolabs.org/​dance). For more info on Anderson's PhD dissertation and the project read below:

The essence of my computational-acoustics dissertation can be boiled down to trying to teach a computer to hear percussion in music like a human. Having a human, Alain Rouvez, teach a robot, Shiny Robot, how to dance seemed like the perfect metaphor.

The dissertation research began with a two-choice, forced-interval experiment in which 29 humans were asked to rate isolated sounds from most to least percussive. The sound characteristic of rise time was found to be the most corrolated with percussion of the characteristics tested. The experiment is repersented in the dance by the first two interactions between Alain and Shiny, during which Shiny expresses his inability to correctly choose the stronger percussion sound.

A rise-time detection algorithm was created based on psychoacoustic models of human hearing. Several parameters of the algorithm, such as number of frequency channels, lowest frequecy of interest, and hair-cell response, are adjustable. The process of tuning these parameters to closely match the algorithm's ranking of sounds to the human ranking was computationally intensive. This tuning process is represented by the last three interactions between Alain and Shiny, as Shiny slowly becomes better at dancing like Alain.

The detection algorithm has many steps, and, after mastering human behavior, Shiny decided to show off a few of them in the dance. Zero padding is represented by the waves traveling along Shiny's arms. Filtering is demonstrated by the artificial stilling of certain appendages during Shiny's frenetic motion. Half-wave rectification is demonstrated by Shiny's left arm only being able to swing back and forth above midline, while his right arm is free. Finally neural firing is represented by Shiny's mule kicks.

The final stage of the dissertation research was to use the detection algorithm with real-world music to discover self-similarity in the percussion patterns. By using auto-correlation analysis, the detection algorithm can be used to time the repetition and near repetition in music percussion. Shiny demonstrates the self-similarity of the music by several final repetitve dance moves, repeating appropriately at the time scale of beats, measures, and phrases.

For the full abstract and dissertation, please see my academic publications page at academic.konfuzo.net/​publications/​publications.php

Category:

Music

Tags:

• Jose
• Lozano
• Magic
• Spoon
• Productions
• Dance
• Your
• PhD
• Video
• Contest
• John
• Anderson
• Mills
• Austin
• Texas