Personal tools

The Science and Technology of Music

Home Activities Workshops and Research Meetings Research Workshops CIRMMT Musical Information Archiving and Retrieval (Axis 3) Workshop

Skip to content. | Skip to navigation

CIRMMT Musical Information Archiving and Retrieval (Axis 3) Workshop

— filed under:

This workshop, with invited guests Elaine Chew and Alexandre François (both from University of Southern California), will take place at McGill University on Friday, February 8th 2008.

  • Research Workshop
When Feb 08, 2008
from 02:00 PM to 07:00 PM
Where A832, New Music Building, McGill University (527 Sherbrooke St. West).
Add event to calendar vCal

This event is free and open to all.


  • 2-3pm: Elaine Chew, University of Southern California: "Music Science and the Mathematical Modeling of Tonality"
  • 3-3:15pm: Coffee break
  • 3:15-4:15pm: Alexandre François, University of Southern California: "A Design Language for Interactive Systems" 
  • 4:15-5pm: General discussion
  • 5-7pm: Reception

Presentation Details

Elaine Chew: ABSTRACT

The surge in interest in the scientific study of music in recent years is spurred on by widespread access to digitally encoded music, from desktop browsers to handheld devices.  A growing group of researchers and scholars are employing mathematical and computational techniques in the scientific analysis and generation of music and expressive performances.  I shall begin with a few words on the field of music science, and some open courseware I have developed on computational modeling of music analysis, performance, and composition/improvisation.

The main part of the talk will focus on mathematical models for recognizing and tracking tonality in music, with live demonstrations of the MuSA.RT real-time tonal analysis and visualization system. Tonality refers to the system by which pitches relate one to another and generate varying degrees of perceived stabilities amongst themselves.  Tonality is central to the understanding of much of the music that we hear.  Questions that pertain to tonality include: when does a phrase sound finished? how does one create expectation, tension and release in a composition? what are some bases for musical humour? how does one manipulate timing to produce the desired coherent groupings or affect?

I shall motivate and introduce the Spiral Array and related algorithms for tonal induction and segmentation.  The Spiral Array is a mathematical model for tonality first introduced in my doctoral dissertation.  It combines Euler's geometric representation for music pitches with concepts from interior point methods in Operations Research.  MuSA.RT shows the Spiral Array representations of current tonal structures as they are computed in real time.  Musical illustrations, with the help of the MuSA.RT system, will provide some answers to the above questions.

The MuSA.RT project is joint work with Alexandre François, who is giving a complementary talk on the underlying software architecture.

Alexandre François: ABSTRACT

Music creation, performance and listening are fundamentally human activities, as music is essentially created by humans for humans. Designers of computer tools that partake in such activities should therefore strive for human-quality interaction.

Computational artifacts remain incapable of naturally interacting with living, biological systems because traditional computation and programming models emulate mathematics' denial of time.  Those of us who strive to design dynamic, adaptive systems that exhibit robust, context-dependent behavior, must consistently attempt to do so with constructs that unnaturally abstract and constrain the very concept of change.  This talk presents a novel, human-centered approach to the design and implementation of interactive software systems, in the form of a design language, Hermes/dl, that bridges this disconnect between mathematical models and natural interaction.

The creation of Hermes/dl builds on the experience gained in creating and using the Software Architecture for Immersipresence (SAI) architectural framework.  A design language comprises of a collection of primitives, a set of organizing principles, and collections of qualifying situations.  The elements of Hermes/dl exist in three interchangeable forms: a human-oriented graphical notation, a proof-oriented graph theoretic formulation, and machine-oriented code middleware.

Hermes/dl's primitives and organizing principles confer modularity and scalability to system designs, and promote the expression of processing logic at the architectural level, prompting the emergence of a rich vocabulary of structural patterns.  Several interactive vision, music and game system designs provide concrete illustrations of these points.  The Crosswinds game system shows the use of the language as a collaboration tool in a class-wide term project.


Document Actions