Predicting the perceived spaciousness of stereophonic music recordings

Andy M. Sarroff, Juan P. Bello

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In a stereophonic music production, music producers seek to impart impressions of one or more virtual spaces upon the recording with two channels of audio. Our goal is to map spaciousness in stereophonic music to objective signal attributes. This is accomplished by building predictive functions by exemplar-based learning. First, spaciousness of recorded stereophonic music is parameterized by three discrete dimensions of perception-the width of the source ensemble, the extent of reverberation, and the extent of immersion. A data set of 50 song excerpts is collected and annotated by humans for each dimension of spaciousness. A verbose feature set is generated on the music recordings and correlation-based feature selection is used to reduce the feature spaces. Exemplar-based support vector regression maps the feature sets to perceived spaciousness. We show that the predictive algorithms perform well on all dimensions and that perceived spaciousness can be successfully mapped to objective attributes of the audio signal.

Original languageEnglish (US)
Title of host publicationProceedings of the 6th Sound and Music Computing Conference, SMC 2009
PublisherSound and music Computing network
Pages83-88
Number of pages6
ISBN (Print)9789899557765
StatePublished - 2009
Event6th Sound and Music Computing Conference, SMC 2009 - Porto, Portugal
Duration: Jul 23 2009Jul 25 2009

Other

Other6th Sound and Music Computing Conference, SMC 2009
CountryPortugal
CityPorto
Period7/23/097/25/09

Fingerprint

Reverberation
Feature extraction

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Sarroff, A. M., & Bello, J. P. (2009). Predicting the perceived spaciousness of stereophonic music recordings. In Proceedings of the 6th Sound and Music Computing Conference, SMC 2009 (pp. 83-88). Sound and music Computing network.

Predicting the perceived spaciousness of stereophonic music recordings. / Sarroff, Andy M.; Bello, Juan P.

Proceedings of the 6th Sound and Music Computing Conference, SMC 2009. Sound and music Computing network, 2009. p. 83-88.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Sarroff, AM & Bello, JP 2009, Predicting the perceived spaciousness of stereophonic music recordings. in Proceedings of the 6th Sound and Music Computing Conference, SMC 2009. Sound and music Computing network, pp. 83-88, 6th Sound and Music Computing Conference, SMC 2009, Porto, Portugal, 7/23/09.
Sarroff AM, Bello JP. Predicting the perceived spaciousness of stereophonic music recordings. In Proceedings of the 6th Sound and Music Computing Conference, SMC 2009. Sound and music Computing network. 2009. p. 83-88
Sarroff, Andy M. ; Bello, Juan P. / Predicting the perceived spaciousness of stereophonic music recordings. Proceedings of the 6th Sound and Music Computing Conference, SMC 2009. Sound and music Computing network, 2009. pp. 83-88
@inproceedings{379a78b0574743fb9b4d8cf8c1ffbb44,
title = "Predicting the perceived spaciousness of stereophonic music recordings",
abstract = "In a stereophonic music production, music producers seek to impart impressions of one or more virtual spaces upon the recording with two channels of audio. Our goal is to map spaciousness in stereophonic music to objective signal attributes. This is accomplished by building predictive functions by exemplar-based learning. First, spaciousness of recorded stereophonic music is parameterized by three discrete dimensions of perception-the width of the source ensemble, the extent of reverberation, and the extent of immersion. A data set of 50 song excerpts is collected and annotated by humans for each dimension of spaciousness. A verbose feature set is generated on the music recordings and correlation-based feature selection is used to reduce the feature spaces. Exemplar-based support vector regression maps the feature sets to perceived spaciousness. We show that the predictive algorithms perform well on all dimensions and that perceived spaciousness can be successfully mapped to objective attributes of the audio signal.",
author = "Sarroff, {Andy M.} and Bello, {Juan P.}",
year = "2009",
language = "English (US)",
isbn = "9789899557765",
pages = "83--88",
booktitle = "Proceedings of the 6th Sound and Music Computing Conference, SMC 2009",
publisher = "Sound and music Computing network",

}

TY - GEN

T1 - Predicting the perceived spaciousness of stereophonic music recordings

AU - Sarroff, Andy M.

AU - Bello, Juan P.

PY - 2009

Y1 - 2009

N2 - In a stereophonic music production, music producers seek to impart impressions of one or more virtual spaces upon the recording with two channels of audio. Our goal is to map spaciousness in stereophonic music to objective signal attributes. This is accomplished by building predictive functions by exemplar-based learning. First, spaciousness of recorded stereophonic music is parameterized by three discrete dimensions of perception-the width of the source ensemble, the extent of reverberation, and the extent of immersion. A data set of 50 song excerpts is collected and annotated by humans for each dimension of spaciousness. A verbose feature set is generated on the music recordings and correlation-based feature selection is used to reduce the feature spaces. Exemplar-based support vector regression maps the feature sets to perceived spaciousness. We show that the predictive algorithms perform well on all dimensions and that perceived spaciousness can be successfully mapped to objective attributes of the audio signal.

AB - In a stereophonic music production, music producers seek to impart impressions of one or more virtual spaces upon the recording with two channels of audio. Our goal is to map spaciousness in stereophonic music to objective signal attributes. This is accomplished by building predictive functions by exemplar-based learning. First, spaciousness of recorded stereophonic music is parameterized by three discrete dimensions of perception-the width of the source ensemble, the extent of reverberation, and the extent of immersion. A data set of 50 song excerpts is collected and annotated by humans for each dimension of spaciousness. A verbose feature set is generated on the music recordings and correlation-based feature selection is used to reduce the feature spaces. Exemplar-based support vector regression maps the feature sets to perceived spaciousness. We show that the predictive algorithms perform well on all dimensions and that perceived spaciousness can be successfully mapped to objective attributes of the audio signal.

UR - http://www.scopus.com/inward/record.url?scp=84905190575&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84905190575&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9789899557765

SP - 83

EP - 88

BT - Proceedings of the 6th Sound and Music Computing Conference, SMC 2009

PB - Sound and music Computing network

ER -