Can listeners hear who is singing? What is the pitch bandwidth of singer discrimination in untrained listeners?

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Objective/Hypothesis: This study sought to determine the pitch bandwidth of singer discrimination based on singer gender, pitch range, and same-different voice category. Study Design: Repeated measures factorial design. Methods: This study used a forced-choice paradigm, where listeners heard two different singers (singer 1 and singer 2) producing /α/ at an identical pitch and an unknown singer (either singer 1 or singer 2) producing /α/ at a different pitch. It was the listener's task to identify which singer (singer 1 or singer 2) was the unknown singer. Two baritones and two tenors were recorded producing /α/ at the pitches C3, E3, G3, B3, D4, and F4. Two sopranos and two mezzo-sopranos were recorded producing /α/ at the pitches C4, E4, G4, B4, D5, and F5. For each group of stimuli, male and female, all possible pairs of singers were constructed for the lowest pitch (C2 or C3, respectively) and highest pitch (F4 or F5, respectively). The unknown singer was varied across the remaining pitches. Results: For between category comparisons, the ability to discriminate singers diminished monotonically with pitch interval, reaching below chance levels between the intervals of the 9th and 11th. For within category comparisons, it was much harder to discriminate singers across pitch, particularly when the voices were female. Conclusion: Timbre is not perceived as invariant across the entire singing range, and the bandwidth of this timbre invariance depends on multiple factors, including pitch range, gender, and same-different voice category.

Original languageEnglish (US)
Pages (from-to)322-329
Number of pages8
JournalJournal of Voice
Volume26
Issue number3
DOIs
StatePublished - May 1 2012

Fingerprint

Singing

All Science Journal Classification (ASJC) codes

  • Otorhinolaryngology
  • Speech and Hearing
  • LPN and LVN

Cite this

Can listeners hear who is singing? What is the pitch bandwidth of singer discrimination in untrained listeners? / Erickson, Mary.

In: Journal of Voice, Vol. 26, No. 3, 01.05.2012, p. 322-329.

Research output: Contribution to journalArticle

@article{c032fb4d0e334f74b24d791b4fa68b62,
title = "Can listeners hear who is singing? What is the pitch bandwidth of singer discrimination in untrained listeners?",
abstract = "Objective/Hypothesis: This study sought to determine the pitch bandwidth of singer discrimination based on singer gender, pitch range, and same-different voice category. Study Design: Repeated measures factorial design. Methods: This study used a forced-choice paradigm, where listeners heard two different singers (singer 1 and singer 2) producing /α/ at an identical pitch and an unknown singer (either singer 1 or singer 2) producing /α/ at a different pitch. It was the listener's task to identify which singer (singer 1 or singer 2) was the unknown singer. Two baritones and two tenors were recorded producing /α/ at the pitches C3, E3, G3, B3, D4, and F4. Two sopranos and two mezzo-sopranos were recorded producing /α/ at the pitches C4, E4, G4, B4, D5, and F5. For each group of stimuli, male and female, all possible pairs of singers were constructed for the lowest pitch (C2 or C3, respectively) and highest pitch (F4 or F5, respectively). The unknown singer was varied across the remaining pitches. Results: For between category comparisons, the ability to discriminate singers diminished monotonically with pitch interval, reaching below chance levels between the intervals of the 9th and 11th. For within category comparisons, it was much harder to discriminate singers across pitch, particularly when the voices were female. Conclusion: Timbre is not perceived as invariant across the entire singing range, and the bandwidth of this timbre invariance depends on multiple factors, including pitch range, gender, and same-different voice category.",
author = "Mary Erickson",
year = "2012",
month = "5",
day = "1",
doi = "10.1016/j.jvoice.2011.06.004",
language = "English (US)",
volume = "26",
pages = "322--329",
journal = "Journal of Voice",
issn = "0892-1997",
publisher = "Mosby Inc.",
number = "3",

}

TY - JOUR

T1 - Can listeners hear who is singing? What is the pitch bandwidth of singer discrimination in untrained listeners?

AU - Erickson, Mary

PY - 2012/5/1

Y1 - 2012/5/1

N2 - Objective/Hypothesis: This study sought to determine the pitch bandwidth of singer discrimination based on singer gender, pitch range, and same-different voice category. Study Design: Repeated measures factorial design. Methods: This study used a forced-choice paradigm, where listeners heard two different singers (singer 1 and singer 2) producing /α/ at an identical pitch and an unknown singer (either singer 1 or singer 2) producing /α/ at a different pitch. It was the listener's task to identify which singer (singer 1 or singer 2) was the unknown singer. Two baritones and two tenors were recorded producing /α/ at the pitches C3, E3, G3, B3, D4, and F4. Two sopranos and two mezzo-sopranos were recorded producing /α/ at the pitches C4, E4, G4, B4, D5, and F5. For each group of stimuli, male and female, all possible pairs of singers were constructed for the lowest pitch (C2 or C3, respectively) and highest pitch (F4 or F5, respectively). The unknown singer was varied across the remaining pitches. Results: For between category comparisons, the ability to discriminate singers diminished monotonically with pitch interval, reaching below chance levels between the intervals of the 9th and 11th. For within category comparisons, it was much harder to discriminate singers across pitch, particularly when the voices were female. Conclusion: Timbre is not perceived as invariant across the entire singing range, and the bandwidth of this timbre invariance depends on multiple factors, including pitch range, gender, and same-different voice category.

AB - Objective/Hypothesis: This study sought to determine the pitch bandwidth of singer discrimination based on singer gender, pitch range, and same-different voice category. Study Design: Repeated measures factorial design. Methods: This study used a forced-choice paradigm, where listeners heard two different singers (singer 1 and singer 2) producing /α/ at an identical pitch and an unknown singer (either singer 1 or singer 2) producing /α/ at a different pitch. It was the listener's task to identify which singer (singer 1 or singer 2) was the unknown singer. Two baritones and two tenors were recorded producing /α/ at the pitches C3, E3, G3, B3, D4, and F4. Two sopranos and two mezzo-sopranos were recorded producing /α/ at the pitches C4, E4, G4, B4, D5, and F5. For each group of stimuli, male and female, all possible pairs of singers were constructed for the lowest pitch (C2 or C3, respectively) and highest pitch (F4 or F5, respectively). The unknown singer was varied across the remaining pitches. Results: For between category comparisons, the ability to discriminate singers diminished monotonically with pitch interval, reaching below chance levels between the intervals of the 9th and 11th. For within category comparisons, it was much harder to discriminate singers across pitch, particularly when the voices were female. Conclusion: Timbre is not perceived as invariant across the entire singing range, and the bandwidth of this timbre invariance depends on multiple factors, including pitch range, gender, and same-different voice category.

UR - http://www.scopus.com/inward/record.url?scp=84860484045&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84860484045&partnerID=8YFLogxK

U2 - 10.1016/j.jvoice.2011.06.004

DO - 10.1016/j.jvoice.2011.06.004

M3 - Article

VL - 26

SP - 322

EP - 329

JO - Journal of Voice

JF - Journal of Voice

SN - 0892-1997

IS - 3

ER -