Lab #4 Pitch Tracking & Intonation

S702 D. Kewley-Port 3/20/02

The purpose of this lab is to use Colea or other pitch extraction procedures to examine intonation patterns for contrastive stress in sentences. Stress is signaled by the acoustic features of F0, duration and intensity. One accepted theory about stress is the Change hypothesis. Perceived stress is a function of the extent of the change of these features, with F0 and duration being the most important in English.

Briefly, in this lab you will produce the sentence, "We were away a year ago" with neutral intonation and then four more times with the stress on 'we', 'were', '(a)way' and 'year'. F0 and duration increments for stressed words and the percent change from the neutral sentence will be calculated. The relative contribution of F0 versus duration to sentence stress will be discussed.

1) Record the sentences in CoolEdit (Fs = 11025Hz) in order to head & tail edit them. Save as a Windows PCM (.wav) file. You should try and record them with the rhythm and duration of the words the same for each repetition so that changes in F0 and intensity are the primary cues for stress.

2) Use Colea (or other pitch tracker) to calculate the F0 intonation contours. Use strategies learned in Homework #6 to calculate the best F0 contour. Save the F0 contours to a file for plotting and analysis.

3) Plot the F0 tracks for the four stressed sentences, overlaid on the neutral one, somewhat like Fig. 6-6, p 235 in Lass. Import the F0 tracks into Excel to do the plotting (or Matlab, but this is much harder). Turn in these four plots, indicating where you made the acoustic measures for the stressed words.

4) Calculate the maximum F0 and duration increments for each stressed word as percent. Assume that increases greater than 10% are above threshold. Determine from your measures if the changes made are perceptible (above threshold), and which acoustic feature was used most consistently to signal stress.

5) Explain the pitch extraction method you chose (in Colea, autocorrelation or cepstrum; see Hess handout, 372-377) in your own words in the methods section.

REFERENCES: From Lass, Kent, p. 213-220 and Lehiste, Chapter 6. for an overview of the concept of the Change hypothesis, see Terken (1991), JASA 89, 1768-1776. Also, see Harris, M. and Umeda, N. (1987). Difference limens for fundamental frequency contours in sentences. Journal of the Acoustical Society of America, 81, 1139-1145.