A time domain description for the pitch strength of iterated rippled noise

William Yost, Roy Patterson, Stanley Sheft

Research output: Contribution to journalArticle

147 Citations (Scopus)

Abstract

Two versions of a cascaded add, attenuate, and delay circuit were used to generate iterated rippled noise (IRN) stimuli. IRN stimuli produce a repetition pitch whose strength relative to the noise can be varied by changing the type of circuit, the attenuation, or the number of iterations in the circuit. Listeners were asked to discriminate between various pairs of IRN stimuli which differed in the type of network used to generate the sounds or the number of iterations (n = 1, 2, 3, 4, 7, and 9). Performance was determined for IRN stimuli generated with delays of 2, 4, and 8 ms and for four bandpass filter conditions (0-2000, 250-2000, 500-2000, and 750-2000 Hz). Some IRN stimuli were extremely difficult to discriminate despite relatively large spectral differences, while other IRN stimuli produced readily discriminable changes in perception, despite small spectral differences. These contrasting results are inconsistent with simple spectral explanations for the perception of IRN stimuli. An explanation based on the first peak of the autocorrelation function of IRN stimuli is consistent with the results. Simulations of the processing performed by the peripheral auditory system (i.e., interval histograms and correlograms) produce results which are consistent with the involvement of these temporal processes in the perception of IRN stimuli.

Original languageEnglish (US)
Pages (from-to)1066-1078
Number of pages13
JournalJournal of the Acoustical Society of America
Volume99
Issue number2
DOIs
StatePublished - Feb 1 1996
Externally publishedYes

Fingerprint

stimuli
iteration
delay circuits
Stimulus
bandpass filters
histograms
autocorrelation
repetition
attenuation
intervals
acoustics
Spectrality

ASJC Scopus subject areas

  • Arts and Humanities (miscellaneous)
  • Acoustics and Ultrasonics

Cite this

A time domain description for the pitch strength of iterated rippled noise. / Yost, William; Patterson, Roy; Sheft, Stanley.

In: Journal of the Acoustical Society of America, Vol. 99, No. 2, 01.02.1996, p. 1066-1078.

Research output: Contribution to journalArticle

@article{e5ffec7e19264be5b94b6dec15f3f95f,
title = "A time domain description for the pitch strength of iterated rippled noise",
abstract = "Two versions of a cascaded add, attenuate, and delay circuit were used to generate iterated rippled noise (IRN) stimuli. IRN stimuli produce a repetition pitch whose strength relative to the noise can be varied by changing the type of circuit, the attenuation, or the number of iterations in the circuit. Listeners were asked to discriminate between various pairs of IRN stimuli which differed in the type of network used to generate the sounds or the number of iterations (n = 1, 2, 3, 4, 7, and 9). Performance was determined for IRN stimuli generated with delays of 2, 4, and 8 ms and for four bandpass filter conditions (0-2000, 250-2000, 500-2000, and 750-2000 Hz). Some IRN stimuli were extremely difficult to discriminate despite relatively large spectral differences, while other IRN stimuli produced readily discriminable changes in perception, despite small spectral differences. These contrasting results are inconsistent with simple spectral explanations for the perception of IRN stimuli. An explanation based on the first peak of the autocorrelation function of IRN stimuli is consistent with the results. Simulations of the processing performed by the peripheral auditory system (i.e., interval histograms and correlograms) produce results which are consistent with the involvement of these temporal processes in the perception of IRN stimuli.",
author = "William Yost and Roy Patterson and Stanley Sheft",
year = "1996",
month = "2",
day = "1",
doi = "10.1121/1.414593",
language = "English (US)",
volume = "99",
pages = "1066--1078",
journal = "Journal of the Acoustical Society of America",
issn = "0001-4966",
publisher = "Acoustical Society of America",
number = "2",

}

TY - JOUR

T1 - A time domain description for the pitch strength of iterated rippled noise

AU - Yost, William

AU - Patterson, Roy

AU - Sheft, Stanley

PY - 1996/2/1

Y1 - 1996/2/1

N2 - Two versions of a cascaded add, attenuate, and delay circuit were used to generate iterated rippled noise (IRN) stimuli. IRN stimuli produce a repetition pitch whose strength relative to the noise can be varied by changing the type of circuit, the attenuation, or the number of iterations in the circuit. Listeners were asked to discriminate between various pairs of IRN stimuli which differed in the type of network used to generate the sounds or the number of iterations (n = 1, 2, 3, 4, 7, and 9). Performance was determined for IRN stimuli generated with delays of 2, 4, and 8 ms and for four bandpass filter conditions (0-2000, 250-2000, 500-2000, and 750-2000 Hz). Some IRN stimuli were extremely difficult to discriminate despite relatively large spectral differences, while other IRN stimuli produced readily discriminable changes in perception, despite small spectral differences. These contrasting results are inconsistent with simple spectral explanations for the perception of IRN stimuli. An explanation based on the first peak of the autocorrelation function of IRN stimuli is consistent with the results. Simulations of the processing performed by the peripheral auditory system (i.e., interval histograms and correlograms) produce results which are consistent with the involvement of these temporal processes in the perception of IRN stimuli.

AB - Two versions of a cascaded add, attenuate, and delay circuit were used to generate iterated rippled noise (IRN) stimuli. IRN stimuli produce a repetition pitch whose strength relative to the noise can be varied by changing the type of circuit, the attenuation, or the number of iterations in the circuit. Listeners were asked to discriminate between various pairs of IRN stimuli which differed in the type of network used to generate the sounds or the number of iterations (n = 1, 2, 3, 4, 7, and 9). Performance was determined for IRN stimuli generated with delays of 2, 4, and 8 ms and for four bandpass filter conditions (0-2000, 250-2000, 500-2000, and 750-2000 Hz). Some IRN stimuli were extremely difficult to discriminate despite relatively large spectral differences, while other IRN stimuli produced readily discriminable changes in perception, despite small spectral differences. These contrasting results are inconsistent with simple spectral explanations for the perception of IRN stimuli. An explanation based on the first peak of the autocorrelation function of IRN stimuli is consistent with the results. Simulations of the processing performed by the peripheral auditory system (i.e., interval histograms and correlograms) produce results which are consistent with the involvement of these temporal processes in the perception of IRN stimuli.

UR - http://www.scopus.com/inward/record.url?scp=0030023201&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030023201&partnerID=8YFLogxK

U2 - 10.1121/1.414593

DO - 10.1121/1.414593

M3 - Article

C2 - 8609290

AN - SCOPUS:0030023201

VL - 99

SP - 1066

EP - 1078

JO - Journal of the Acoustical Society of America

JF - Journal of the Acoustical Society of America

SN - 0001-4966

IS - 2

ER -