VF_2007b - The Male-Female Coalescence Model as Applied to...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: The Male-Female Coalescence Model as Applied to Nonsingers and Trained Singers Trained James Harnsberger W.S. Brown, Jr. Howard Rothman Rahul Shrivastav 36th Annual Symposium: Care of the 36th Professional Voice Professional May 31, 2007 Introduction Introduction Question 1: Do male and female voices Question coalesce in mean fundamental frequency (F0) with age? – Coalescence: Higher in older males and lower in Higher older females • • Cross-sectional: Cross-sectional: – Female: Saxman & Burk (1967); Morgan Rastatter (1986); Brown, Morris, Mitchell (1989) – Male and Female: Krook (1988) - Swedish; Brown, Morris, Hollien, Howell (1991); Higgins & Saxman (1991) Longitudinal: Longitudinal: – Female: Russell, Penny, Pemberton (1995); de Pinto & Hollien (1982) Introduction Introduction • Question 1: Do male and female voices Question coalesce in mean F0 with age? – No change • Males: Wilcox & Horii (1980); Ramig & Ringel (1983); Wilcox Orlikoff (1990) Orlikoff – Other • Decoster and Debruyne (2000): Male f0 increased Decoster significantly with age in a longitudinal study (Dutch) significantly Introduction Introduction • Coalescence appears in Coalescence the transition from middle to old age middle • Note: Increase in f0 in Note: females with very old age age • Coalescence involves Coalescence modest changes in fundamental frequency fundamental Figure from Baken (2005) Introduction Introduction Question 2: Does vocal training influence any Question changes in average fundamental frequency due to age? due – Training influences perceived age (Harnsberger, Training Brown, Rothman, Morris, Shrivastav 2006) Brown, • Old singers have significantly lower perceived age than Old old nonsingers, regardless of gender old – No coalescence observed in male and female No singers (Brown, Morris, Hollien, and Howell, 1991) singers • Caveat: Small sample sizes for trained and untrained Caveat: groups groups Introduction Introduction • Experiment 1 - Age Estimation 70 – 65 Percei ved Age (Years) Perceived Age Perceived Results 60 Singer 55 Nonsinger 50 45 40 Female Male Gender Group Training: F (1,21) Training: = 184, p < 0.01 184, • • Gender: n.s. Training*Gender: Training*Gender: n.s. Coalescence in Singers (Brown et al (1991) (1991) 240 220 Mean f0 (Hz) 200 224 218 214 206 192 195 191 188 180 175 160 140 120 137 126 118 100 138 136 127 121 119 100 80 Young Middle Old Age Group Female Untrained Male Untrained Female Sopranos Male Tenors Female Altos Male Baritones Present Study Present • Acoustic analysis of large database of trained Acoustic singers and nonsingers singers • Mean fundamental frequency • Does coalescence occur? Does – If so, does it affect singers and nonsingers If equally? • If not, does a lack of coalescence account for differences If between singers and nonsingers in perceived age? between Methods Methods • Subjects (180 nonsingers, 80 singers) 45 39 40 35 26 24 25 21 19 20 15 13 10 10 9 8 5 Female Young Male Female Male Middle Subject Pool Female Trained Untrained Trained Untrained Trained Untrained Trained Untrained Trained Untrained Trained 0 Untrained N 30 32 30 29 Male Old Methods Methods • Stimulus materials: Second sentence of the Stimulus rainbow passage rainbow – “The rainbow is a division of white light into many The beautiful colors” beautiful • Substudy: Varying stimulus materials in Substudy: nonsingers nonsingers – – – – Passages (Rainbow, Grandfather) Sentences (16 SPIN sentences) Sentences Sustained vowels: [a], [i], [u] Diadodes: “pataka”, “shupupi” Diadodes: Results - Nonsingers Results 200 190 193 180 Mean f0 (Hz) 170 170 163 160 150 140 130 120 110 127 116 115 100 Young Middle Age Group Female Untrained Male Untrained Old Results - Singers Results 200 190 180 185 183 178 Mean f0 (Hz) 170 160 150 140 130 120 110 120 122 107 100 Young Middle Age Group Female Trained Male Trained Old Results Results • Statistical Analysis (ANOVA) – – – – – – – Age: n.s. Gender: F(1,248) = 377.3, p < 0.01 Training: n.s. Age * Gender: F(2,248) = 9.7, p < 0.01 Age * Training: F(2,248) = 2.9, p = 0 .06 Training * Gender: n.s. Age*Gender*Training: n.s. • Post hoc tests – Males: Young, Middle < Old – Females: Young > Middle, Old Results - Substudy Results Coalesence Across Materials Mean fundamental frequency (Hz) 220 200 Female Diadodes Female Grandfather Female Rainbow Female SPIN Female Vowel Male Diadodes Male Grandfather Male Rainbow Male SPIN Male Vowel 180 160 140 120 100 Young Middle Age Group Old Discussion Discussion • Coalescence observed in this large database Coalescence in both singers and nonsingers in – Male voices gradually increase in mean Male fundamental frequency across the three age groups groups – Female voices decrease from young to middle age Discussion Discussion • Coalescence Coalescence – Trends in mean F0 were consistent across all Trends types of speech materials for nonsingers types – Note: F0 variability differed across materials in Note: terms of demonstrating an age effect (Old vs. Young) Young) • Female: Vowel > Grandfather, Rainbow > SPIN > Diadodes • Male: Diadodes > Rainbow > SPIN > Grandfather > Vowel Discussion Discussion • Singers show the same coalescence trend, Singers despite the fact that their voices are perceived to age more slowly for both genders to – Note: Differs from Brown et al 1991, who used Note: smaller groups smaller • Coalescence remains a modest effect – 10% increase in males – 16% decrease in females Future Work Future • Acoustic analysis of new cues for age in Acoustic singers and nonsingers singers – Voice quality • – – Breathiness Vowel formants F0 variability • Cycle to cycle ...
View Full Document

Ask a homework question - tutors are online