Overview

The Vowel Demonstrator provides an introduction to the acoustic consequences of articulatory manipulations involved in the production of vowels.

The Vowel Demonstrator introduces the notions of source-filter theory, that the sound from a periodic glottal source is filtered by the vocal tract, imposing spectral peaks that determine which vowel we perceive. Selecting a location in the “vowel quadrilateral” leads to: 1) a display of a schematic representation of the vocal tract configuration corresponding to that vowel; 2) displays of the source spectrum, the vocal tract filter function, and the output spectrum for that vowel; and 3) audio of the synthesized vowel. Changing the vocal tract leads to easily observable changes in the filter function, its peaks, and the output spectrum. Thus, the acoustic consequences of varying tongue position and lip rounding can be easily demonstrated.

The Vowel Demonstrator is a useful tool for teaching courses in acoustic phonetics and speech science. With the aid of a vowel synthesizer, it provides visual and aural demonstrations that emphasize the ideas of source-filter theory, linking vocal tract configuration to the vowel quadrilateral, with the identities of vowels being associated with formant frequencies (filter peaks).

Hardware Requirements

  • Intel Core i3 CPU @ 3 GHz or better
  • 4 GB RAM
  • 300 MB available hard-disk space
  • 1600 x 900 monitor resolution or higher
  • Internet connection (for activation/installation)

Software Requirements

  • Microsoft Windows® 7/8/10
  • Microsoft .Net Framework 4.5 (included)
  • Java 7 v.1.7.0_55 or higher (free download)

Documentation

The Vowel Demonstrator illustrates the basics of vowel production using a speech synthesizer to model vowel production.

Vowel Demonstrator Source Spectrum

This figure shows the source spectrum of the vowel. The acoustic source for vowels is vibration of the glottis. Glottal vibrations result from muscle actions that both 1) push air between the vocal folds, tending to separate them; and 2) tense the vocal folds, tending to pull them back together. Because the vocal folds are elastic, these forces repeatedly force the vocal folds gradually apart and then snap them back together more rapidly. Since the open/close cycles are regular and produced at approximately equal time intervals of T0, the resulting sound is periodic, with an amplitude spectrum composed of harmonics (multiples) of the fundamental frequency f0 = 1/T0. The shape of the spectral envelope (that is, the imaginary curve connecting the peaks of the harmonics) is determined by the exact shape (gradual opening, rapid closure) of the open/close cycle.

Vowel Demonstrator Filter

This figure shows the filtering effect of the current vocal tract configuration. The sound produced by the glottis is filtered as it passes through the vocal tract because the vocal tract has resonances which amplify some frequencies more than others. Each configuration of the vocal tract acts as a different acoustic filter because it has different resonances. The resonances of the vocal tract result in more amplification and higher output at some frequencies and less amplification and lower output at frequencies. The bands of higher output are called formants. Vowels are characterized primarily by the frequencies of the first three formants, labeled F1, F2, and F3 (in ascending order). The relative amplitude of the formants is a secondary factor that affects vowel quality.

Click or drag on the “meters” (red or blue) below the graph of the filter to move the corresponding cursor (the red or blue vertical line) to one of the peaks in the graph. When the cursor is at a peak in the filter, it measures a formant frequency. The formant frequency is displayed in the meter, e.g., “1525 Hz”. The gain of the filter at the formant frequency is displayed at the top of the graph, in dB (e.g., “22 dB”). The corresponding cursors (red or blue vertical lines) in the graph of the source spectrum (to the left) and the output spectrum (to the right) display the corresponding amplitudes at the same frequency.

This figure shows the spectrum of the output sound. The output sound, the sound that would be recorded in front of the mouth, is the result of the filter acting on the source spectrum. It reflects both the downward tilt of the source spectrum and the resonances of the filter created by the vocal tract. At any given frequency, the amplitude of the output sound (in decibels, or dB) is equal to the amplitude of the glottal source plus the amplification (also called “gain”) from the filter.

Vowel Demonstrator Fundamental

Click or drag on the slider’s handle to adjust the fundamental frequency. The synthesizer will play the glottal source sound while you are adjusting the fundamental. Note that the spacing between harmonics changes as fundamental frequency changes, because the harmonics are at frequencies which are multiples of the fundamental frequency (2 x f0, 3 x f0, 4 x f0, etc.) The shape of the source spectrum is determined by the open/close cycle, which is set in the speech synthesizer. Note also that the envelope of the source spectrum by and large does not change when the vocal tract configuration changes.

Vowel Demonstrator Vocal Tract

This figure shows a sagittal section of a vocal tract, including the lips, tongue, jaw, and teeth. Different vowels are produced by different configurations of these articulators. Changing the shape and size of the hollow spaces (cavities) in the vocal tract thus changes the resonances of the vocal tract. The most important factor is the position of the tongue in the oral cavity. Tongue position is described by how high, and how far back, the tongue is in the mouth. In the Vowel Demonstrator, vocal tract configuration is controlled by positioning the cursor in the vowel quadrilateral.

Vowel Demonstrator Lips Rounded

Another important factor in vowel production is lip rounding. With the tongue in a given position, the character of a vowel can be changed markedly by rounding or unrounding the lips.
Click on the “lip rounding” checkbox to the right to change the lip posture. ition of the lips.

Vowel Demonstrator Vocal Quadrilateral

The vowel quadrilateral is a schematic roughly indicating tongue position for the production of various rounded and unrounded vowels. The left side of the quadrilateral indicates vowels in which the tongue is positioned towards the front of the mouth; the right side indicates vowels with the tongue towards the back. The top of the vowel quadrilateral indicates vowels with the tongue raised close to the roof of the mouth; the bottom of the vowel quadrilateral indicates low tongue and jaw positions that produce a more open vocal tract.

Click or drag on the vowel quadrilateral to change the vocal tract configuration and listen to the corresponding synthetic vowel.

Vowel Demonstrator IPA Cardinals

Click on the “IPA cardinals” checkbox to the right to see the IPA symbols for rounded and unrounded vowels on the vowel quadrilateral.

FAQ

Q. Do I need Java installed?

A. Yes. You must have at least Java 7 (specifically, JRE 1.7.0_55); a 32-bit version on 32-bit Windows® 7 and a 64-bit version for 64-bit Windows® 7 and all versions of Windows® 8 and Windows® 10. You can obtain a current version of Java directly from Oracle at www.java.com.

Q. I get an error when I try to download and install Java.

A. First, consult your institution’s tech support. There may be firewalls or filters that would prevent the download. Also, you generally must be an Administrator in order to install new software.

We have observed an error in some browsers in which the Java download page attempts to download something similar to “JavaSetup8u31.com&AuthParam= …” or “jre-8u31-windows-i586.com&AuthParam=…”
WORKAROUND:

Chose “Save As” when the browser asks whether or not you want to download the file. This option may be in a drop-down menu under browser “Save” button. Keep the base filename (“JavaSetup8u31” or “jre-8u31-windows-i586”) and change the file type to “exe”. The result will be something like “JavaSetup8u31.exe” or “jre-8u31-windows-i586.exe”. You may also have to change the file type in the Save dialog box to “All Files”.

If the Java installer does not automatically start after the download finishes, navigate to your “Downloads” folder and double-click it to start installation.

Q. When I try to run the Vowel Demonstrator, I get an error message “No audio device available. Please make sure that headphones or speakers are connected to your computer.”

A. First, right-click on the loudspeaker icon in the System Tray and Open the Volume Mixer to check if audio for the Vowel Demonstrator is muted. If it is, unmute it. Also make sure the volume turned up.

Second, make sure that there is either a loudspeaker or headphones connected to your computer’s audio output jack.

Q. I get an error message like “server not available” or “server not reachable” when I try to Activate Vowel Demonstrator.

A. WORKAROUND:
Close the dialog with the error message, then open a browser window and go to www.licensespot.com. After the page loads, click the “Activate” button in the Activator again.

Q. What speech synthesizer is used in Vowel Demonstrator?

A. A version of Sensimetrics’ “LLSYN” formant synthesizer is used in this application. LLSYN contains a number of parameters for tailoring the glottal source (e.g. the open quotient) and includes various acoustic factors that are sometimes neglected in simpler formant synthesizers (e.g., coupling to tracheal poles and zeros during the open phase of the glottal oscillation).