The Corposition of MusAIc: from MIRsic to CritAIcs

Wednesday 15 February 2017, 4.00PM

Speaker(s): Dr Nick Collins (Durham University)

Artificially intelligent musAIcians can gain much from the existence and automatic analysis of large audio file corpora. The study of databases of musical audio, often on the order of listening days or weeks of examples, has been much investigated in music information retrieval (MIR) in recent decades. Associated "MIRsic" exploits machine listening and machine learning machinery over large music collections in the creation of new works. The neologism "corposition" denotes composition with corpora, which has precedents in plunderphonics, concatenative synthesis, stylistic algorithmic composition and sampling libraries.

In this presentation, the general backdrop and trends of such work will be examined, as well as detailing specific compositions based in associated techniques, such as the chronological sampling electroacoustic study 'A History of Electronic Music' or the interactive Schubert/contemporary music discriminator 'Schubot' for piano and electronics. Application in automatic music critics ("critAIcs") will be considered, including an artificial judging trio for a dubstep remix competition, and in-concert critic systems meant to provoke performers with their influences.

Nick Collins is Reader in Composition at Durham University. His research interests include live computer music, musical artificial intelligence, and computational musicology, and he is a frequent international performer as composer-programmer-pianist or codiscian, from algoraves to electronic chamber music. Many research papers and much code and music is available from

Location: Sally Baldwin D Block D003