Generative Music Systems for Live Performance
Andrew R. Brown, Toby Gifford, and Rene Wooller
Queensland University of Technology, Brisbane , Australia.
{a.brown, t.gifford, r.wooller}@qut.edu.au
Music improvisation continues to be an intriguing area for computational creativity.
In this paper we will outline two software systems designed for live music
performance, the LEMu (live electronic music) system and the JamBot (improvisatory
accompaniment agent). Both systems undertake an analysis of human created music,
generate complementary new music, are designed for interactive use in live
performance, and have been tested in numerous live settings. These systems have
some degree of creative autonomy, however, we are especially interested in the
creative potential of the systems interacting with human performers.
The LEMu software generates transitional material between scores provided in
MIDI format. The LEMu software uses an evolutionary approach to generated
materials that provide an appropriate path between musical targets [1]. This musical
morphing process is controlled during performance by an interactive nodal graph that
allows the performer to select the morphing source and target as well as transition
speed and parameters. Implementations included the MorphTable [2] where users
manipulate blocks on a large surface to control musical morphing transitions. This
design suits social interaction and is particularly suited to use by inexperienced users.
The JamBot [3] listens to an audio stream and plays along. It consists of rhythmic
and harmonic analysis algorithms that build a dynamic model of the music being
performed. This model holds multiple probable representations at one time in the
Chimera Architecture [4] which can be interpreted in various ways by a generative
music algorithm that adds accompaniment in real time.
These systems have been designed using a research method we have come to call
Generation in Context, that relies on iterations of aesthetic reflection on the generated
outcomes to inform the processes of enquiry [5].
<references_biblio/>
References
1. Brown, A.R., R. Wooller, and E.R. Miranda, Interactive Evolutionary Composition: Musical
morphing as a compositional strategy, in A-Life for Music: Music and Computer Models of
Living Systems, E.R. Miranda, Editor. 2010, A-R Editions: Middleton, WI, (in press).
2. Brown, A. R., R. Wooller, and K. Thomas, The Morph Table: A collaborative interface for
musical interaction, in Trans: Boundaries/Permeability/Reification. Australasian Computer
Music Conference, A. Riddel and A. Thorogood, Editors. 2007, ACMA:Canberra. p. 34–39.
3. Gifford, T., JamBot. 2008: Brisbane. http://www.dr.offig.com/
4. Gifford, T. and A. R. Brown, Do Androids Dream of Electric Chimera?, in Improvise: The
Australasian Computer Music Conference 2009, A. Sorensen and P. McIlwain, Editors.
2009, ACMA: Brisbane.
5. Brown, A. R., T. Gifford, E. Narmour and R. Davidson, Generation in Context: An
Exploratory Method for Musical Enquiry in The Second International Conference on Music
Communication Science, C. Stevens, et al., Editors. 2009, HCSNet: Sydney. p. 7-10.
290