candles image

Steven Hargreaves

centre for digital music
queen mary, university of london

Research Interests

Multitrack Audio Metadata Capture

Current Music Information Retrieval (MIR) research focuses mostly on fully mixed music. Consequently, a significant proportion of any MIR task is the isolation and identification of the relevant and interesting components of a musical signal - melodic, dynamic or rhythmic information, for example. The proliferation of Digital Audio Workstations (DAWs) though means that a large, untapped, body of source-separated material exists from which a vast amount of accurate audio metadata may be derived.

The central aim of my PhD research was to investigate methods of exploiting multitrack data for early, accurate, metadata capture using a combination of signal processing, machine learning and logical inferencing techniques.


After graduating from Manchester Metropolitan University in 1993 with a degree in Mathematics, I spent several years working in a record shop before returning to academia, firstly to study Computer Science (MSc, Edinburgh, 1998), then Audio Acoustics (MSc, Salford, 2004), and most recently Music Metadata Capture in the Studio from Audio and Symbolic Data (PhD, Queen Mary University of London, 2014). In between those degrees, I worked as a Java web-development consultant and latterly an Audio Engineer, developing software and measurement systems for tuning the parameters of mobile phone audio processors. In parallel to all those activities I've played in several bands and tried my hand at recording and production. I am currently working as a research assistant at the Centre for Digital Music, Queen Mary University of London.

last updated 13th June 2015