The VowelWorm: Real-time Mapping and Visualisation of Sung Vowels in Music

Publication Type:

Conference Paper


SMC Conference 2011 (2011)


This paper presents an approach to predicting vowel quality in vocal music performances, based on common acoustic features (mainly MFCCs). Rather than performing classification, we use linear regression to project spoken or sung vowels into a continuous articulatory space: the IPA Vowel Chart. We introduce a real-time on-line visualisation tool, the Vowel Worm, which builds upon the resulting models and displays the evolution of sung vowels over time in an intuitive manner. The concepts presented in this work can be used for artistic purposes and music teaching.

smc2011_submission_167.pdf4.1 MB
SMC paper: