Please use this identifier to cite or link to this item:
https://hdl.handle.net/2440/139580
Type: | Thesis |
Title: | A Methodology For Analysing Improvised Jazz: A Computer-Aided Approach |
Author: | Blackwell, David James |
Issue Date: | 2023 |
School/Discipline: | Elder Conservatorium of Music |
Abstract: | This musicological research employs a computer-aided statistical approach to the analysis of improvised jazz. The main aim of this study is to develop a new methodology for systematically analysing improvisational style, with this research completed in two parts. This is done first through an analysis of Grant Green’s improvisational style, based on transcriptions of forty improvisations between 1960 and 1965. Green (1935–1979) was a prolific but underrated jazz guitarist, and was the unofficial house guitarist for Blue Note Records between 1960 and 1965. This research aims to explore, analyse, and explain Green’s improvisational style with reference to his use of pitch, rhythm, micro, and macro features. Secondly, the results of this analysis are used to inform performer classifications and comparative analysis between Grant Green, John Coltrane, Miles Davis, and Charlie Parker. Tree based machine learning algorithms are utilised to complete the performer classification tasks, with the comparative analysis based upon the features found to classify the performers. This research built upon previous work from the Jazzomat Research Project (2012–2017), based out of the University of Music Franz Liszt Weimar. This research uses methods and software developed by the Jazzomat Research Project to transcribe and extract the data from Green’s solos, with the data for the other three performers in the comparative analysis coming from their Weimar Jazz Database. The analyses, and training and evaluation of the machine learning classifiers, were undertaken in the R programming language. The results of this study found that Green conformed to many of the improvisational conventions of the time, with these results confirming the validity of the developed methodology. Findings from the classification task found that the C5.0 classifier was the most efficient and performant when classifying the improvisers. The results of this research contribute to the field of computational musicology and the analysis of improvised jazz. The methodology developed through this research will allow future investigations to thoroughly explore the improvisational style of other musicians. |
Advisor: | Harrald, Luke Corn, Aaron |
Dissertation Note: | Thesis (Ph.D.) -- University of Adelaide, Elder Conservatorium of Music, 2023 |
Keywords: | Jazz guitar musicology computational musicology computer-aided musicology statistics Grant Green machine learning jazz theory Jazzomat jazz analysis music analysis |
Provenance: | This electronic version is made publicly available by the University of Adelaide in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. This thesis may incorporate third party material which has been used by the author pursuant to Fair Dealing exceptions. If you are the owner of any included third party copyright material you wish to be removed from this electronic version, please complete the take down form located at: http://www.adelaide.edu.au/legals |
Appears in Collections: | Research Theses |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
Blackwell2023_PhD.pdf | 70.84 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.