OCAD University Open Research Repository

Vocal Synthetics: Designing for an Adaptable Singing Synthesizer

Clarke, Liam (2021) Vocal Synthetics: Designing for an Adaptable Singing Synthesizer. Masters thesis, OCAD University.

Item Type: Thesis
Creators: Clarke, Liam
Abstract:

Technological music tools such as digital audio workstations and electronic music instruments have enabled musicians without formal training to create music that is heard by millions of people. The automation by software and hardware can create compelling productions without limitations from performance ability. However, the automation of vocals is particularly difficult because beyond pitch and timbre, the vocalization of language requires additional parameters for control. As the production of a vocal synthesizer and its vocal palettes is complex, the current market sees these difficulties represented through products that have limited voices and do not adapt to vocal trends. This project demonstrates a tool that allows producers to use a simple typing interface for the input of words, allowing the output to be integrated and controlled by modern digital audio workstations. Using a machine learning solution, the tool is not dependent on large stores of audio data once a model is trained and since it contains a simple method to create new voices, it can keep up with evolving musical trends and vocal styles. The aim is to bring the human voice into the realm of digital music production enabling a music maker to include a large range of vocal styles within their production tool set. This paper outlines the design and development of the tool and culminates in a piece of music that illustrates the value of applying design thinking research strategies to an artistic and technical challenge.

Date: 1 January 2021
Divisions: Graduate Studies > Digital Futures
Date Deposited: 11 May 2021 20:35
Last Modified: 20 Dec 2021 20:31
URI: https://openresearch.ocadu.ca/id/eprint/3448

Actions (login required)

Edit View Edit View