Upload
bernadette-patrick
View
217
Download
0
Tags:
Embed Size (px)
Citation preview
Hoarse meeting in Liverpool April 22, 2005
Subglottal pressure and NAQ variation in Classically Trained Baritone Singers
Eva Björkner*†, Johan Sundberg†, Paavo Alku *Laboratory of Acoustics and Audio Signal Processing, Helsinki University of Technology, Finland
†Department of Speech Music Hearing, KTH, Stockholm, Sweden
Air => Subglottal pressure
(Ps)
Oscillation =>pulsating air flow => voice source
Sound => vowels and consonants
Voice production
TEAC Multi channel digital recorder
Flow (Rothenberg mask)
Audio
Oral pressure
Inverse Filtering
DeCap – Svante Granqvist
Flow glottogram Flow glottogram parametersparameters (Time-based & amplitude-based)
Time [s]
0,002 0,004 0,006 0,008 0,01 0,012
0,002 0,004 0,006 0,008 0,01 0,012
Time [s]
Time [T0]
Derivative [MFDR]
Ûp-t-p
Peak-to-peak pulse amplitude
Flo
w
Deri
vati
ve
Flow
Negative peak of the differentiated flow
Information about vocal loudness and phonation type are reflected in the changes of the glottal closing phase.
The first studies using parameterization of the glottal flow based on amplitude domain measurements was made by Fant & Lin in 1988.
In 1994 Fant et al. introduced the effective declination time and presented a time-domain measure by computing the ratio between two amplitude values, the AC-flow and the derivative of the differentiated flow, eg., the maximum flow declination rate MFDR
Glottal closing phase
NAQ, the normalized AQNAQ, the normalized AQ
AQ
T0Alku P, Bäckström T, Vilkman E. (2002)
In parallel with Fant´s studies,In parallel with Fant´s studies, Alku & Vilkman Alku & Vilkman introduced the introduced the Amplitude Quotient AQAmplitude Quotient AQ in in studies 1996studies 1996
Ûp-t-p
MFDR
Maximum flow declination rate
which normalizes the AQ values with respect to the duration of the fundamental period T0.
Alku et al. found that the AQ parameter systematically reflected changes in phonation mode and that AQ differed between sexes.
What kind of information about voice production
can NAQ give?
It has so far been used in studies about :
Speech
intensity
phonation type
vocal loading
emotional expressions
voice quality
Singing
singing styles
register
Subglottal pressure and NAQ variation in Classically Trained Baritone Singers
Eva Björkner*†, Johan Sundberg†, Paavo Alku
five Swedish professional baritone singers
international opera carriers
age range 29-65 years
A sung diminuendo at a constant pitch while repeating the syllable [pae:]
Sung at three F0 located at approximately 25%, 50% and 75% of their professional pitch range
Ten equally spaced Ps-values were selected from each singers total Ps range.
Au
dio
Pre
ssu
re
Mean Ps for three F0
0
5
10
15
20
25
30
35
40
45
1 2 3 4 5 6 7 8 9 10Pressure number
Pre
ssure
[cm
H2O
]
~139 Hz~196 Hz~277 Hz
Ps means for the 5
singers
Highly structured data
Fundamental frequency and pressure are strongly correlated
Singer 3
y = 2,1124x - 5,9948
R2 = 0,982
0
10
20
30
40
50
60
0 5 10 15 20 25
Mean pressures ~139 Hz
Mea
n pr
essu
res
~27
7 H
z
Singer 4
y = 2,567x - 3,1855R2 = 0,9911
0
10
20
30
40
50
60
0 5 10 15 20 25
Mean pressures ~139 Hz
Mean p
ress
ure
s ~
277 H
z
Singer 2
y = 2,3068x + 0,7386
R2 = 0,9853
0
10
20
30
40
50
60
0 5 10 15 20 25
Mean pressures ~139 Hz
Mean p
ress
ure
s ~
277 H
z
Singer 5
y = 1,9824x - 6,2238R2 = 0,9657
0
10
20
30
40
50
60
0 5 10 15 20 25
Mean pressures ~139 Hz
Mean p
ress
ure
s ~
277 H
z
Singer 1
y = 2,2338x - 2,0326
R2 = 0,9947
0
10
20
30
40
50
60
0 5 10 15 20 25
Mean pressures ~139 Hz
Mean p
ress
ure
s ~
277 H
z
Pressure differences between octaves
0
500
1000
1500
2000
2500
3000
0 5 10 15 20 25 30 35 40
Mean Pressure [cmH2O]
Mea
n M
FDR
[l/
s2]
HIGH
LOW
Means across the 5 singers
MFDR increases with increasing Ps
For the same Ps the low F0 shows higher MFDR-values
=> due to the longer period time
NAQ => Information about phonation type
NAQ decreases with
increasing MFDR
NAQ decreases with
increasing Ps
NAQ differs with F0NAQ differs with F0
NAQ and Pressure
0,00
0,05
0,10
0,15
0,20
0,25
0,30
0,35
0,40
0 10 20 30 40 50 60Pressure [cmH2O]
NAQ
NAQ & MFDR
0,00
0,05
0,10
0,15
0,20
0,25
0,30
0,35
0,40
0 1000 2000 3000 4000MFDR
NA
Q
Open symbol=high F0
Filled symbol= low F0
Higher NAQ-values for higher F0…
-Does that mean that these professional singers change phonation type with increasing F0??!!
Why these differences when we have normalized?!
AQ and Ps
0
0,0005
0,001
0,0015
0,002
0,0025
0,003
0 10 20 30 40 50 60
Pressure [cmH2O]
AQ
AQ & MFDR
0
0,0005
0,001
0,0015
0,002
0,0025
0,003
0 1000 2000 3000 4000MFDR
AQ
AAQQ
Open symbol=high F0
Filled symbol= low F0
What happened?
An effect of the singers´ skill to keep the same phonation type independently of Ps and F0?
AQ should give a steady value if phonation type is kept
# 1 AQ
0
0.0005
0.001
0.0015
0.002
0.0025
0 2 4 6PSEN [cmH2O]
AQ
# 2 AQ
0
0.0005
0.001
0.0015
0.002
0.0025
0 2 4 6
Psen [cmH2O]AQ
#4 AQ
0
0,0005
0,001
0,0015
0,002
0,0025
0 2 4 6 8 10 12Psen [cmH2O]
AQ
# 5 AQ
0
0,0005
0,001
0,0015
0,002
0,0025
0 2 4 6 8
Psen [cmH2O]
AQ
# 3 AQ
0
0,0005
0,001
0,0015
0,002
0,0025
0 2 4 6
Psen [cmH2O]
AQ
NAQ or AQ
Speaker do not have large changes in F0, but change phonation mode
Singers have large changes in F0, but do not change phonation mode => AQ…
Conclusions
The five singer’s Ps data were highly structured
Approximately a doubling of Ps for a doubling of F0
Does AQ more accurately reflect phonation mode than NAQ
?
For the same Ps the low F0 showed higher MFDR-values
For a given Ps increase MFDR increased more at low F0
The End