Show simple item record

dc.contributor.supervisorMiranda, Eduardo
dc.contributor.authorMetters, Laurence
dc.contributor.otherFaculty of Arts, Humanities and Businessen_US
dc.date.accessioned2021-04-12T11:22:16Z
dc.date.available2021-04-12T11:22:16Z
dc.date.issued2021
dc.identifier10474421en_US
dc.identifier.urihttp://hdl.handle.net/10026.1/17033
dc.description.abstract

This thesis presents an investigation into the uses of machine learning and artificial intelligence for electronic sound synthesis, specifically the creation of new synthesised sounds for composition and research. Using the Magenta Labs Neural Synthesizer (NSynth), a synthesis tool that uses deep neural networks to generate new sounds based on data input from electronic synthesizers, this research project aimed to produce a system where bespoke synthesizers could be used to produce interesting sound combinations consisting of approaches to electronic sound synthesis that would not conventionally be used together. Combinations of different approaches to electronic sound synthesis produced interesting results when choices were made based on sonic characteristics of individual synthesisers, such as the plucked timbres of a Karplus-Strong synthesizer and the smooth extended notes of a frequency modulation synthesizer. In this thesis, contextual information into both electronic sound synthesis including justification for the use of each synthesis method as well as an investigation into Artificial Intelligence(AI) and machine learning techniques relevant to the use of the NSynth has also been carried out with the intention of producing an informed and researched final product in the form of a composition. The summary of this research project culminated in a final composition utilising the sound samples produced by the NSynth, arranged into a piece inspired by computer music research compositions including John Chowning’s ‘Stria’ and ‘Switched on Bach’ by Wendy Carlos. The synthesizers in this research project were produced in Max and designed with specific sonic qualities of the instruments in mind, with versatility to produce more samples following the same process for further research and application by other composers inspired by the work. The resulting composition included a number of interesting sounds with plenty of variation in sonic qualities that resembled computer music composition as well as standard western composition, demonstrating the versatility of the concept of using AI and bespoke synthesisers to create new and interesting sounds.

en_US
dc.language.isoen
dc.publisherUniversity of Plymouth
dc.rightsCC0 1.0 Universal*
dc.rights.urihttp://creativecommons.org/publicdomain/zero/1.0/*
dc.subjectArtificial Intelligenceen_US
dc.subjectAIen_US
dc.subjectComputer Musicen_US
dc.subjectMusicen_US
dc.subjectNeural networken_US
dc.subjectElectronic Sound Synthesisen_US
dc.subject.classificationResMen_US
dc.titleAN INVESTIGATION INTO THE USES OF MACHINE LEARNING FOR ELECTRONIC SOUND SYNTHESISen_US
dc.typeThesis
plymouth.versionpublishableen_US
dc.identifier.doihttp://dx.doi.org/10.24382/1196
dc.rights.embargoperiodNo embargoen_US
dc.type.qualificationDoctorateen_US
rioxxterms.versionNA


Files in this item

Thumbnail
Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record

CC0 1.0 Universal
Except where otherwise noted, this item's license is described as CC0 1.0 Universal

All items in PEARL are protected by copyright law.
Author manuscripts deposited to comply with open access mandates are made available in accordance with publisher policies. Please cite only the published version using the details provided on the item record or document. In the absence of an open licence (e.g. Creative Commons), permissions for further reuse of content should be sought from the publisher or author.
Theme by 
Atmire NV