Drum Pattern Humanization Using a Recursive Bayesian Framework
×
Cite This
Citation & Abstract
R. Stables, C. Athwal, and R. Cade, "Drum Pattern Humanization Using a Recursive Bayesian Framework," Paper 8763, (2012 October.). doi:
R. Stables, C. Athwal, and R. Cade, "Drum Pattern Humanization Using a Recursive Bayesian Framework," Paper 8763, (2012 October.). doi:
Abstract: In this study we discuss some of the limitations of Gaussian humanization and consider ways in which the articulation patterns exhibited by percussionists can be emulated using a probabilistic model. Prior and likelihood functions are derived from a dataset of professional drummers to create a series of empirical distributions. These are then used to independently modulate the onset locations and amplitudes of a quantized sequence, using a recursive Bayesian framework. Finally, we evaluate the performance of the model against sequences created with a Gaussian humanizer and sequences created with a Hidden Markov Model (HMM) using paired listening tests. We are able to demonstrate that probabilistic models perform better than instantaneous Gaussian models, when evaluated using a 4/4 rock beat at 120 bpm.
@article{stables2012drum,
author={stables, ryan and athwal, cham and cade, rob},
journal={journal of the audio engineering society},
title={drum pattern humanization using a recursive bayesian framework},
year={2012},
volume={},
number={},
pages={},
doi={},
month={october},}
@article{stables2012drum,
author={stables, ryan and athwal, cham and cade, rob},
journal={journal of the audio engineering society},
title={drum pattern humanization using a recursive bayesian framework},
year={2012},
volume={},
number={},
pages={},
doi={},
month={october},
abstract={in this study we discuss some of the limitations of gaussian humanization and consider ways in which the articulation patterns exhibited by percussionists can be emulated using a probabilistic model. prior and likelihood functions are derived from a dataset of professional drummers to create a series of empirical distributions. these are then used to independently modulate the onset locations and amplitudes of a quantized sequence, using a recursive bayesian framework. finally, we evaluate the performance of the model against sequences created with a gaussian humanizer and sequences created with a hidden markov model (hmm) using paired listening tests. we are able to demonstrate that probabilistic models perform better than instantaneous gaussian models, when evaluated using a 4/4 rock beat at 120 bpm.},}
TY - paper
TI - Drum Pattern Humanization Using a Recursive Bayesian Framework
SP -
EP -
AU - Stables, Ryan
AU - Athwal, Cham
AU - Cade, Rob
PY - 2012
JO - Journal of the Audio Engineering Society
IS -
VO -
VL -
Y1 - October 2012
TY - paper
TI - Drum Pattern Humanization Using a Recursive Bayesian Framework
SP -
EP -
AU - Stables, Ryan
AU - Athwal, Cham
AU - Cade, Rob
PY - 2012
JO - Journal of the Audio Engineering Society
IS -
VO -
VL -
Y1 - October 2012
AB - In this study we discuss some of the limitations of Gaussian humanization and consider ways in which the articulation patterns exhibited by percussionists can be emulated using a probabilistic model. Prior and likelihood functions are derived from a dataset of professional drummers to create a series of empirical distributions. These are then used to independently modulate the onset locations and amplitudes of a quantized sequence, using a recursive Bayesian framework. Finally, we evaluate the performance of the model against sequences created with a Gaussian humanizer and sequences created with a Hidden Markov Model (HMM) using paired listening tests. We are able to demonstrate that probabilistic models perform better than instantaneous Gaussian models, when evaluated using a 4/4 rock beat at 120 bpm.
In this study we discuss some of the limitations of Gaussian humanization and consider ways in which the articulation patterns exhibited by percussionists can be emulated using a probabilistic model. Prior and likelihood functions are derived from a dataset of professional drummers to create a series of empirical distributions. These are then used to independently modulate the onset locations and amplitudes of a quantized sequence, using a recursive Bayesian framework. Finally, we evaluate the performance of the model against sequences created with a Gaussian humanizer and sequences created with a Hidden Markov Model (HMM) using paired listening tests. We are able to demonstrate that probabilistic models perform better than instantaneous Gaussian models, when evaluated using a 4/4 rock beat at 120 bpm.
Authors:
Stables, Ryan; Athwal, Cham; Cade, Rob
Affiliation:
Birmingham City University, Birmingham, UK
AES Convention:
133 (October 2012)
Paper Number:
8763
Publication Date:
October 25, 2012Import into BibTeX
Subject:
Sound Analysis and Synthesis
Permalink:
http://www.aes.org/e-lib/browse.cfm?elib=16505