A perceptual spectral difference model for binaural signals

Calum Armstrong; Thomas McKenzie; Damian Murphy; Gavin Kearney

A perceptual spectral difference model for binaural signals

Calum Armstrong, Thomas McKenzie, Damian Murphy, Gavin Kearney

Research output: Contribution to conference › Paper › peer-review

Abstract

This paper presents a perception based model for calculating the difference between two binaural signals, to more accurately represent the perceptual relevance of spectral differences. A basic spectral difference calculation, the difference between the fast Fourier transforms (FFTs) of two audio signals, is not an accurate metric for human perception as the auditory system differs greatly in sensitivity depending on relative amplitude, frequency and temporal aspects. The presented model is evaluated through objective measures and comparison to the results of a previously published listening test.

Original language	English
Publication status	Published - 7 Oct 2018
Event	145th Audio Engineering Society International Convention, AES 2018 - New York, United States Duration: 18 Oct 2018 → 21 Oct 2018

Conference

Conference	145th Audio Engineering Society International Convention, AES 2018
Country/Territory	United States
City	New York
Period	18/10/18 → 21/10/18

Bibliographical note

Publisher Copyright:
© 2018 KASHYAP.

Cite this

@conference{0067d6d6ef61458eacdde12d1f2f907d,

title = "A perceptual spectral difference model for binaural signals",

abstract = "This paper presents a perception based model for calculating the difference between two binaural signals, to more accurately represent the perceptual relevance of spectral differences. A basic spectral difference calculation, the difference between the fast Fourier transforms (FFTs) of two audio signals, is not an accurate metric for human perception as the auditory system differs greatly in sensitivity depending on relative amplitude, frequency and temporal aspects. The presented model is evaluated through objective measures and comparison to the results of a previously published listening test.",

author = "Calum Armstrong and Thomas McKenzie and Damian Murphy and Gavin Kearney",

note = "Publisher Copyright: {\textcopyright} 2018 KASHYAP.; 145th Audio Engineering Society International Convention, AES 2018 ; Conference date: 18-10-2018 Through 21-10-2018",

year = "2018",

month = oct,

day = "7",

language = "English",

}

TY - CONF

T1 - A perceptual spectral difference model for binaural signals

AU - Armstrong, Calum

AU - McKenzie, Thomas

AU - Murphy, Damian

AU - Kearney, Gavin

PY - 2018/10/7

Y1 - 2018/10/7

N2 - This paper presents a perception based model for calculating the difference between two binaural signals, to more accurately represent the perceptual relevance of spectral differences. A basic spectral difference calculation, the difference between the fast Fourier transforms (FFTs) of two audio signals, is not an accurate metric for human perception as the auditory system differs greatly in sensitivity depending on relative amplitude, frequency and temporal aspects. The presented model is evaluated through objective measures and comparison to the results of a previously published listening test.

AB - This paper presents a perception based model for calculating the difference between two binaural signals, to more accurately represent the perceptual relevance of spectral differences. A basic spectral difference calculation, the difference between the fast Fourier transforms (FFTs) of two audio signals, is not an accurate metric for human perception as the auditory system differs greatly in sensitivity depending on relative amplitude, frequency and temporal aspects. The presented model is evaluated through objective measures and comparison to the results of a previously published listening test.

UR - http://www.scopus.com/inward/record.url?scp=85055032174&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85055032174

T2 - 145th Audio Engineering Society International Convention, AES 2018

Y2 - 18 October 2018 through 21 October 2018

ER -

A perceptual spectral difference model for binaural signals

Abstract

Conference

Bibliographical note

Other files and links

Cite this