SLIDE 1 On the relationship between input parameters in the two-mass vocal-fold model with acoustical coupling and signal parameters of the glottal flow
- A. Van Hirtum1, I. Lopez2, M. Hirschberg2, X. Pelorson1
1ICP, INPG, Grenoble, France 2TUE, Eindhoven, The Netherlands
SLIDE 2
Introduction voiced sound glottal volume air flow (Ug) analysis synthesis production generative model physical model
SLIDE 3 Introduction generative function two-mass model glottal volume air flow (Ug) generative model physical model
acoustical signal physiology & fluid flow simplification of reality
SLIDE 4 Objective
To assess the sensitivity
with acoustical coupling to the input-parameters
SLIDE 5 input parameters (I) (physiology, fluid flow, intrinsic)
physical two-mass model (Lous)
glottal volume flow Ug
generative function (LF)
glottal volume flow Ug signal parameters (S) (acoustical signal, intrinsic)
Objective
voice quality parameters (time)
SLIDE 6 Two-mass vocal fold model
physiology mechanics fluid flow vocal fold sub-glottal (trachea h0) supra-glottal (vocal tract h3) y x Psub mtot
K1, ξ1 K2, ξ2 hc
SLIDE 7
Vocal tract configuration
/a/ /i/
SLIDE 8
Signal time-parameters /a/ (S)
Q0=t0/T0 Qs=tp/tn Qr=Ap/An U0 , D0 , F0 , Te , Ta , Ag (hmin) Qm=U0/D0
SLIDE 9
Relative sensitivity index
ref ref var ref ref var
)/I I (I )/S S (S β[-] − − =
SLIDE 10
Two-mass model input parameter reference values (I)
x1/x2 h3 ξc Kc mtot h0 Lg 12 mm 0.0002 mm Lin, Lout 0.06 1000 Pa Psub /a/, /i/ 0.01 mm y1, y2 1.1 0.1 ξ1, ξ2 20 N/m 40 N/m K1, K2 0.2 g 0.1 g m1, m2 3 mm 10 mm Lw
SLIDE 11 0.0005 0.0047 0.0007 0.0015 0.29 0.0002 2.90 3.18 0.70 104
/i/
0.0006 0.014 0.0010 0.0022 0.24 0.0002 1.37 1.78 0.69 108
/a/
hmin [m] Ta [s] Te [s] Qm [s] D0 [m3/s2] U0 [m3/s] Qr [-] Qs [-] Q0 [-] F0 [Hz]
Reference signal parameters from Ug (dUg)
SLIDE 12
/a/ /i/ >600 Sensitivity F0 Qs U0 Qm Ta Q0 Qr D0 Te hmin
SLIDE 13
SLIDE 14
/a/ /i/
/a/ /i/ ≠
F0 Qs U0 Qm Ta Q0 Qr D0 Te hmin
SLIDE 15
/a/ /i/ F0 Qs U0 Qm Ta Q0 Qr D0 Te hmin
SLIDE 16 /a/ /i/
) β(K β(K)
c
>
F0 Qs U0 Qm Ta Q0 Qr D0 Te hmi
SLIDE 17 Conclusion
- sensitivity of the input-parameters of physical two-mass
model on signal-parameters is quantified in terms of vocal quality parameters and 4 Liljencrants-Fant input parameters
- influence of vocal tract shape on the sensitivity
SLIDE 18
Introduction voiced sound glottal volume air flow (Ug) analysis synthesis production
SLIDE 19
Introduction analysis synthesis production glottal volume air flow (Ug) generative model physical model
