Data and recordings associated with the article:
Hillenbrand, J., Getty, L. A., Clark, M. J., & Wheeler, K. (1995). Acoustic characteristics of American English vowels. The Journal of the Acoustical Society of America, 97(5), Article 5. https://doi.org/10.1121/1.411872
Hosted with permission from Jim Hillenbrand.
Below is the original readme file with some slight formatting changes. This is found in the h95-alldata.zip file. All files described are in the same zip file.
character 1: m=man, w=woman, b=boy, g=girl characters 2-3: talker number characters 4-5: vowel (ae="had", ah="hod" (the vowel in "cot"), aw="hawed", eh="head", er="heard", ey="hayed", ih="hid", iy="heed", oa="hoed" (/o/ as in "boat"), oo="hood",uh="hud", uw="who'd)
Note that there are a few talker numbers that are not used (e.g., m05, w18, g03).
men.zip: Sound files for men compressed as a pkzip/Winzip file. 16 kHz linear PCM, Windows .wav format.
women.zip: Ditto for women.
kids.zip: Ditto for kids.
vowdata.dat: Acoustic measurements (file format explained at the head of the file). Formants sampled at: (a) steady state, (b) 20% of vowel duration, (c) 50% of vowel duration, and (d) 80% of vowel duration. Text file. See top of file for an explanation of the file format.
bigdata.dat: Same as 'vowdata.dat', but with finer sampling of the formant contour (every 10% of vowel duration). Text file. See top of file for an explanation of the file format.
vowdata.ds: Descriptive statistics for acoustic measurements. Text file.
iddata.dat: Labeling data from 20 listeners. Text file.
misid.dat: Subset of 'iddata.dat' listing misidentified tokens only; i.e., tokens for which the majority vote was not the vowel intended by the talker. Text file.
timedata.dat: Vowel start, stop, and steady-state times for each signal. Text file. See top of file for an explanation of the file format.