2. “Connecting the dots” How do articulatory processes “map” onto acoustic processes?

3. Stevens and House (1955) Model assumes  No coupling with Nasal cavity trachea & pulmonary system

4. Stevens and House (1955) Model parameters  Distance of major constriction from glottis (d 0 )  Radius of major constriction (r 0 )  Area (A) and length (l) of lip constriction A/l conductivity index Figure 1.

5. Comparing model to real vocal tract

6. Stevens and House (1955) Figure 2.

7. Key Goal of Study  Evaluate the effect of systematically changing each of these three “vocal tract” parameters on F1-F3 frequency

8. Formant Frequency (KHz) Point of Constriction (d 0 ) (cm from glottis) F1 F2 F3 Figure 3.

9. Formant Frequency (KHz) Point of Constriction (d 0 ) (cm from glottis) F1 F2 F3 Figure 3.

11. Point of constriction A/l NOTE  Single intersection between F1 & F2 in most cases Figure5.

12. Point of constriction A/l Figure 5.

13. Point of constriction A/l Figure 7.

14. General Observations ∆ d 0 = ∆ V front & V back ↑ d 0 = ↓ V front = ↑ F2 ↑ d 0 = ↑ V back = ↓ F1

15. General Observations ↓ r 0 = ↓ F1 ↑ r 0 = ↑ F1 When d 0 ↑ (anterior) ↓ r 0 = ↓ V front = ↑ F2 ↑ lip rounding = ↓ A/l = ↓ F1 & F2

18. r0r0 d0d0 - + - +

21. 20 “normalizing” formant values

22. Clinical Example

23. 22 Acoustic variables related to the perception of vowel quality  F1 and F2  Other formants (i.e. F3)  Fundamental frequency (F 0 )  Duration  Spectral dynamics i.e. formant change over time

24. How helpful is F1 & F2? Data SourceHuman ListenersPattern Classifier Peterson & Barney (1952) 94.4 %74.9 % Hillenbrand et al. (1995) 95.2 %68.2 % From Hillenbrand & Gayvert (1993)

25. How does adding more variables improve pattern classifier success?  F1, F2 + F3 80-85 %  F1, F2 + F 0 80-85 %  F1, F2 + F3 + F 0 89-90 %

26. How about Duration? Nearby vowels have different durations

28. What about Duration?

29. Some examples

30. What about formant variation?

32. Naturally spoken /hAd/ Synthesized, preserving original formant contours Synthesized with flattened formants What about formant variation?

33. Conclusion: Spectral change patterns do matter. What about formant variation?

34. What do we conclude?

35. Sinewave Speech Demonstration Sinewave speech examples (from HINT sentence intelligibility test):

36. Selected issues that are not resolved  What do listener’s use? Specific formants vs. spectrum envelope  What is the “planning space” used by speakers? Articulatory Acoustic Auditory