Abstract
| Original language | English |
|---|---|
| Title of host publication | Innovation in Music 2015 |
| Editors | R. Hepworth-Sawyer, J. Hodgson, R. Toutson |
| Volume | 2 |
| Publication status | Published - 7 Jun 2015 |
Publication series
| Name | KES Transactions on Innovation in Music |
|---|
Fingerprint
Keywords
- compositional algorithm
- Hidden Markov Model
- software-based performance
- Max for Live platform
Cite this
}
Interactive algorithmic composition. / Averell, Edward; Knox, Don.
Innovation in Music 2015. ed. / R. Hepworth-Sawyer; J. Hodgson; R. Toutson. Vol. 2 2015. (KES Transactions on Innovation in Music).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Interactive algorithmic composition
AU - Averell, Edward
AU - Knox, Don
PY - 2015/6/7
Y1 - 2015/6/7
N2 - The research described in this paper gathers insights from music cognition research and applies them to the creation of a software-based performance and compositional tool which has the ability to improvise a musical sequence in accompaniment to user input. The software was developed using Cycling 74’s Max, with incorporation into Ableton Live through the ‘Max for Live’ platform. The program employs a compositional algorithm which follows models of melodic expectation to react and adapt to musical input, creating melodic accompaniment in real-time. A rule set of four features is employed, controlling the selection of notes in relation to harmonic compliance, metric salience, post-skip reversal and intervallic distance incorporating late phrase declination. The system also allows the user to identify 'favourite' musical passages created by the system. From this input, a Hidden Markov Model (HMM) approach is used to generate new sequences based on the preferences of the musician. This paper describes the design and operation of the interactive algorithmic system in detail. Objective analysis of the system has shown it to consistently output phrases which correspond with models of melodic expectation, generating a suitable and consistent musical accompaniment which adapts to user input.
AB - The research described in this paper gathers insights from music cognition research and applies them to the creation of a software-based performance and compositional tool which has the ability to improvise a musical sequence in accompaniment to user input. The software was developed using Cycling 74’s Max, with incorporation into Ableton Live through the ‘Max for Live’ platform. The program employs a compositional algorithm which follows models of melodic expectation to react and adapt to musical input, creating melodic accompaniment in real-time. A rule set of four features is employed, controlling the selection of notes in relation to harmonic compliance, metric salience, post-skip reversal and intervallic distance incorporating late phrase declination. The system also allows the user to identify 'favourite' musical passages created by the system. From this input, a Hidden Markov Model (HMM) approach is used to generate new sequences based on the preferences of the musician. This paper describes the design and operation of the interactive algorithmic system in detail. Objective analysis of the system has shown it to consistently output phrases which correspond with models of melodic expectation, generating a suitable and consistent musical accompaniment which adapts to user input.
KW - compositional algorithm
KW - Hidden Markov Model
KW - software-based performance
KW - Max for Live platform
M3 - Conference contribution
SN - 9781911108047
VL - 2
T3 - KES Transactions on Innovation in Music
BT - Innovation in Music 2015
A2 - Hepworth-Sawyer, R.
A2 - Hodgson, J.
A2 - Toutson, R.
ER -