Geometric Analysis of Musical Audio Data

Project Summary

In this work, we turn musical audio time series data into shapes for various tasks in music matching and musical structure understanding. 

Themes and Categories
Year
Contact
Christopher Tralie
Electrical and Computer Engineering
chris.tralie@gmail.com

In particular, we use sliding window representations of chunks of audio to create high dimensional time-ordered point clouds, and we extract information by analyzing the geometry of these clouds.  We have shown, for example, that sequences of these shapes can be used to identify two different versions of the same song, or "cover songs."  We have also shown that both local and global musical properties can be expressed in geometric language.  For instance, hip hop is very "wiggly" while classical music is very "smooth."  Choruses and verses tend to live in distinct clusters connected by paths, and "bridges," or detours to a different musical idea, show up as large loops.

Related People

Related Projects

A team of students led by History Professor Cecilia Márquez will use census data to understand the long history of Latinxs in the U.S. South. Despite a growing focus of historians and social scientists on the historical and contemporary Latinx South, there has not yet been a thorough data analysis of the historical presence of Latinxs in the South. The Data+ team will search the U.S. Federal Census, immigration records, and marriage records to determine the location of Latinxs in the U.S. South over the course of the late nineteenth and early twentieth centuries. This work will provide an invaluable data set to help us understand the long southern history of Latinxs. 

Project Lead: Cecilia Márquez

Are the concepts of a “consumer” and of a “consumer society” modern ideas? Is greed good, as Michael Douglas’s Gordon Gekko in the 1987 movie Wall Street claimed, or is it a destructive sin?

A team of students led by Dr. Astrid Giugni (Duke, English and ISS) and Dr. Jessica Hines (Brimingham-Southern College, English) will address the question of how to trace concepts that slowly developed alongside changing economic and social realities.

 

We will track a set of related terms (such as consumer, greed, speculation, profit) in order to begin assessing how the ethical, political, and economic language of goods-consumption changed around the Protestant Reformation and the rise of the market economy. Using large databases-- EEBO (Proquest), ECCO (Gale), HathiTrust, and TEAMS (University of Rochester)—that contain scans and machine-readable Medieval and Early Modern texts, the group will track and analyze pamphlets, sermons, satires, and images to understand how the ethical discourse of consumerism changed over time.  

Project Leads: Astrid Giugni, Jessica Hines

Project Manager: Chris Huebner

A team of students led by Humanities Unbounded Fellow Eva Michelle Wheeler will explore how culturally-bound language in African-American literature and film is rendered for international audiences and will map where and into which languages these translations are occurring. Students will use a reference dataset to build and annotate a translation corpus, explore the lexical choices and translation strategies employed by translators, and conduct a macro-level analysis of the geographic and linguistic spread of these types of translations. The results of this project will bring a quantitative dimension to what has largely been a qualitative analysis and will contribute to ongoing academic conversations about language, race, and globalization.  

Project Lead: Eva Wheeler