ELVIS—the Electronic Locator of Vertical Interval Successions.

Five-Minute Introduction

From the Montréal ELVIS team.

Go »

ISMIR 2014

Materials related to our presentation at the International Society for Music Information Retrieval conference in Taipei.

Go »

Source Code

Our software is available under free-and-open licences.

Go »

Sonification

Turning music data into sound.

Go »

Quick Description

The Basic Problem

For centuries, music scholars have researched counterpoint as one of the fundamental elements of Western music. With the advent of computer-driven music analysis, most analytic efforts have been based on melody or harmony—two equally-important musical elements. Now is the time for counterpoint to catch up.

Equipped with Schubert and Cumming's “contrpuntal modules”, Myke Cuthbert's music21 library for Python, the Perl/Humdrum software collection, and a growing set of online repositories of symbolic musical data, the ELVIS team has set out to bring an empirical basis to music theoretic conjecturing about counterpoint.

Funding

ELVIS started in earnest when we received one of 14 Digging into Data Challenge awards in the 2011 competition. Funded jointly by several international funding organizations (like SSHRC from Canada and JISC from the United Kingdom), Digging into Data asks participating teams to use “big data” techniques in their humanities research projecs. For the ELVIS team, this means coming to terms with 600 years of recorded Western music in the form of vertical interval successions.

The Research Team

As with all Digging into Data projects, the ELVIS team comprises several international teams. We have team members working at:

Our Work

Each team has made their own contribution to the ELVIS project's overall project. The McGill team has collected a repository of symbolic musical scores with carefully-selected metadata, developed a framework for music analysis programs, and will soon deploy a Web-based application for analyzing contrapuntal modules.

The MIT team has been working hard to furnish new functionality for music21 that eases analysis of large musical collections, increasing stability and efficiency along the way.

The Yale team has furnished the database with a sizeable collection of 18th- and 19th-century music while performing important statistical work that may call into question aspects of our received knowledge of harmony.

Finally, the Aberdeen team has been answering questions about stylistic differences between similar composers by combining symbolic research with perception-and-cognition testing.

Of course, all teams have presented at professional conferences, and are preparing numerous scholarly publications.

ELVIS Computational Music Analysis