COMING SOON! PQDT Open is getting a new home!

ProQuest Open Access Dissertations & Theses will remain freely available as part of a new and enhanced search experience at

Questions? Please refer to this FAQ.

Dissertation/Thesis Abstract

Motion capture to build a foundation for a computer -controlled instrument by study of classical guitar performance
by Norton, Jonathan Carey, Ph.D., Stanford University, 2008, 151; 3332997
Abstract (Summary)

Using an experimental analytic approach, the specific questions asked are: (1) How exactly is the guitar sound produced by the player? (2) Can the gestures required for each sound be identified? (3) Can motion capture depict the mechanics of how the instrument is played precisely enough to be analyzed? (4) Can the motion captured articulations be categorized in terms of degrees of freedom? The results from these questions can be used to determine the optimal degrees of freedom necessary to build a computer-controlled representation of the classical guitar in the future and may contribute to the basic study of similar instruments.

First, the performer/instrument interface, i.e., how the performer produces sound with the instrument, must be understood. This is explained through the evolution of the classical guitar and its effect on how the performer's seating, left-hand and right-hand positions and support of the instrument developed to its current state.

With an understanding of the performer/instrument interface the next question is: What is a musical gesture? A musical gesture can be defined as the grouping of articulations (sub-gestures) for expressive purposes. A musical gesture not only includes articulations but also includes expressive ancillary movements. In this manner, articulations form the building blocks that are used to create phrasing and musical gestures and represent the way in which guitar tones are: (1) attacked, (2) sustained, and (3) released. There exists in this sense an important distinction between musical gestures and articulations. Viewing articulations as building blocks allows each articulation to be studied discretely and addressed as a separate entity without having to contend with how to connect one articulation to another, possibly overlapping as co-articulations. If all of the classical guitar articulations can be defined, grouped and captured, then, in theory, all classical guitar music should be able to be reproduced.

In the study, thirty markers are placed on the specific areas of interest on the body. In the case of the classical guitarist, these sites are the fingers, thumb and wrist joints of both hands, the elbows and the shoulders. The guitarist repeats each articulation four times at three separate dynamic levels. Singular value decomposition (SVD), a type of multi-dimensional scaling analysis used for finding patterns in highly dimensional data, and a novel method of auto-correlation and peak detection are then applied to the data. The result is scalable data reduction which yields the separate dimensional contribution of the markers for each individual articulation. Four methods of threshold analysis are then applied to the marker contributions of each articulation, each method providing a different perspective. Lastly, the concept of regions of interest (ROI) is put forth as a way to gain insight into the dimensional data.

Indexing (document details)
School: Stanford University
School Location: United States -- California
Source: DAI-A 69/10, Dissertation Abstracts International
Subjects: Music, Music education, Robotics
Keywords: Articulations, Classical guitar, Computer-controlled instrument, Guitar, Motion capture, Musical articulations, Musical gestures, Region of interest
Publication Number: 3332997
ISBN: 978-0-549-85130-1
Copyright © 2021 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy