Toward a computational model of perceived spaciousness in recorded music

Andy M. Sarroff, Juan P. Bello

Research output: Contribution to journalArticle

Abstract

A computational model of perceived spaciousness in recorded musical signals, inspired by research in music information retrieval (MIR), is presented and evaluated experimentally. First three dimensions of spaciousness are selected for computational modeling: width of source ensemble, extent of reverberation, and extent of immersion. Next human subject responses to stereophonic music along these dimensions were collected. Then audio features from this data set of music were extracted and finally three exemplar-based machine learning models for estimating the three dimensions of spaciousness were trained and tested. The worst predictor was found to perform at least 32% better than a baseline predictor. These results are important to the music and audio engineering communities, as computational models for spaciousness in music may present new, perceptually meaningful methods of analysis and processing for the spatial characteristics of recorded music.

Original languageEnglish (US)
Pages (from-to)498-513
Number of pages16
JournalAES: Journal of the Audio Engineering Society
Volume59
Issue number7-8
StatePublished - Jul 2011

Fingerprint

Reverberation
Information retrieval
Learning systems
Processing
Recorded music
Computational Model
Music
Predictors
Ensemble
Computational Modeling
Machine Learning
Music Information Retrieval
Exemplar-based
Learning Model
Human Subjects
Immersion

ASJC Scopus subject areas

  • Music
  • Engineering(all)

Cite this

Toward a computational model of perceived spaciousness in recorded music. / Sarroff, Andy M.; Bello, Juan P.

In: AES: Journal of the Audio Engineering Society, Vol. 59, No. 7-8, 07.2011, p. 498-513.

Research output: Contribution to journalArticle

@article{d4f89634d35b40d8b8c28686dd7783ef,
title = "Toward a computational model of perceived spaciousness in recorded music",
abstract = "A computational model of perceived spaciousness in recorded musical signals, inspired by research in music information retrieval (MIR), is presented and evaluated experimentally. First three dimensions of spaciousness are selected for computational modeling: width of source ensemble, extent of reverberation, and extent of immersion. Next human subject responses to stereophonic music along these dimensions were collected. Then audio features from this data set of music were extracted and finally three exemplar-based machine learning models for estimating the three dimensions of spaciousness were trained and tested. The worst predictor was found to perform at least 32{\%} better than a baseline predictor. These results are important to the music and audio engineering communities, as computational models for spaciousness in music may present new, perceptually meaningful methods of analysis and processing for the spatial characteristics of recorded music.",
author = "Sarroff, {Andy M.} and Bello, {Juan P.}",
year = "2011",
month = "7",
language = "English (US)",
volume = "59",
pages = "498--513",
journal = "AES: Journal of the Audio Engineering Society",
issn = "1549-4950",
publisher = "Audio Engineering Society",
number = "7-8",

}

TY - JOUR

T1 - Toward a computational model of perceived spaciousness in recorded music

AU - Sarroff, Andy M.

AU - Bello, Juan P.

PY - 2011/7

Y1 - 2011/7

N2 - A computational model of perceived spaciousness in recorded musical signals, inspired by research in music information retrieval (MIR), is presented and evaluated experimentally. First three dimensions of spaciousness are selected for computational modeling: width of source ensemble, extent of reverberation, and extent of immersion. Next human subject responses to stereophonic music along these dimensions were collected. Then audio features from this data set of music were extracted and finally three exemplar-based machine learning models for estimating the three dimensions of spaciousness were trained and tested. The worst predictor was found to perform at least 32% better than a baseline predictor. These results are important to the music and audio engineering communities, as computational models for spaciousness in music may present new, perceptually meaningful methods of analysis and processing for the spatial characteristics of recorded music.

AB - A computational model of perceived spaciousness in recorded musical signals, inspired by research in music information retrieval (MIR), is presented and evaluated experimentally. First three dimensions of spaciousness are selected for computational modeling: width of source ensemble, extent of reverberation, and extent of immersion. Next human subject responses to stereophonic music along these dimensions were collected. Then audio features from this data set of music were extracted and finally three exemplar-based machine learning models for estimating the three dimensions of spaciousness were trained and tested. The worst predictor was found to perform at least 32% better than a baseline predictor. These results are important to the music and audio engineering communities, as computational models for spaciousness in music may present new, perceptually meaningful methods of analysis and processing for the spatial characteristics of recorded music.

UR - http://www.scopus.com/inward/record.url?scp=80052576914&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80052576914&partnerID=8YFLogxK

M3 - Article

VL - 59

SP - 498

EP - 513

JO - AES: Journal of the Audio Engineering Society

JF - AES: Journal of the Audio Engineering Society

SN - 1549-4950

IS - 7-8

ER -