Double vowelsDouble vowels
Pairs of synthetic vowels that start and stop at the same time (so-called 'double vowels') have been used to demonstrate that listeners can use information about the fundamental frequency (F0) of sounds to separate them. Specifically, listeners are better able to identity the constituents of a double vowel when the two vowels have different F0s, relative to the condition in which they have the same F0.
This application allows you to experiment with the set of five vowels used in the studies of Assmann and Summerfield (1990). Many modelling studies have also used this vowel set (e.g. Meddis and Hewitt, 1992). Each of the five vowels is synthesized on six F0s, corresponding to differences of between 0 and 4 semitones from a 100 Hz fundamental:
Semitone difference
Fundamental Frequency (Hz)
Period (ms)
zero
100.00
10.0
quarter
101.45
9.86
half
102.93
9.72
one
105.95
9.44
two
112.25
8.90
four
126.00
7.94
Use the menus on the upper left of the screen (1) to select the constituents of a vowel pair. The amplitude ratio of the two vowels can be adjusted using the slider (2). The waveform of the vowel mixture is shown on the upper right of the screen (3), together with buttons which allow you to hear the mixture, and each individual vowel component (4).
The lower panels of the screen show the rectified basilar membrane response to the vowel mixture (5) and the correlogram (6). Below the correlogram is the summary (or pooled) autocorrelation function (7), which will tend to exhibit peaks of activity at the periods of the two vowels in the mixture. If the checkbox Show pitch periods (9) is activated, red lines in the summary function will mark the periods of the two vowels.
On slower machines, the update of the basilar membrane and correlogram displays may be sluggish. You can turn off automatic updating of these displays by deactivating the checkbox labelled Automatically update (9). When automatic updating is disabled, you can force a recomputation of the displays by pressing the Recompute button (8).
1. Do you always see a peak in the summary autocorrelation function at the period of each vowel? Do you always see a peak at the period of at least one of the vowels?
2. What does your answer to (1) suggest about possible strategies for segregating double vowels?
P. F. Assmann and Q. Summerfield (1990) Modelling the perception of concurrent vowels: vowels with different fundamental frequencies. JASA 88 (2), 680-697.
R. Meddis and M. J. Hewitt (1992) Modelling the identification of concurrent vowels with different fundamental frequencies. JASA, 91(1), 233-245.
See also the demonstrations for vowel segregation using neural oscillations (vowelSeg).
Produced by: Guy J. Brown
Release date: June 22 1998
Permissions: This demonstration may be used and modified freely by anyone. It may be distributed in unmodified form.