Talking Heads: Speech Production

Measuring and Modeling Speech Production

The Acoustic Theory of Speech Production: the source-filter
model
Acoustic speech output in humans and many nonhuman species is commonly
considered to result from a combination of a source of sound energy (e.g. the
larynx) modulated by a transfer (filter) function determined by the shape of
the supralaryngeal vocal tract. This combination results in a shaped spectrum
with broadband energy peaks. This model is often referred to as the "source-
filter theory of speech production" and stems from the experiments of
Johannes Mller (1848) in which a functional theory of phonation was tested
by blowing air through larynges excised from human cadavers. "Mller ...
noticed that the sound that came directly from the larynx differed from the
sounds of human speech. Speechlike quality could be achieved only when he
placed over the vibrating cords a tube whose length was roughly equal to the
length of the airways that normally intervene between the larynx and a
persons lips. The sound then resembled the vowel [uh], the first vowel in the
word about ..." (from Lieberman, 1984). In this model the source of acoustic
energy is at the larynx the supralaryngeal vocal tract serves as a variable
acoustic filter whose shape determines the phonetic quality of the sound
(Fant, 1960).
When the larynx serves as a source of sound energy, voiced sounds are
produced by a repeating sequence of events. First, the vocal cords are brought
together (adduction), temporarily blocking the flow of air from the lungs and
leading to increased subglottal pressure. When the subglottal pressure
becomes greater than the resistance offered by the vocal folds, they open
again. The folds then close rapidly due to a combination of factors, including
their elasticity, laryngeal muscle tension, and the Bernoulli effect. If the
process is maintained by a steady supply of pressurized air, the vocal cords
will continue to open and close in a quasiperiodic fashion. As they open and
close, puffs of air flow through the glottal opening. The frequency of these
pulses determines the fundamental frequency (F) of the laryngeal source
and contributes to the perceived pitch of the produced sound. An example of
the spectrum of the result of such glottal air flow is plotted at the top left of
Figure 2. Note that there is energy at the fundamental frequency (F = 100
Hz) and at the harmonics of the fundamental, and that the amplitude of the
harmonics falls off gradually. The bottom left panel shows the comparable
case for a fundamental frequency of 200 Hz. The rate at which the vocal folds
open and close during phonation can be varied in a number of ways and is
determined by the tension of the laryngeal muscles and the air pressure
generated by the lungs. The shape of the spectrum is determined by details of
the opening and closing movement, and is partly independent of fundamental
frequency. In normal speech fundamental frequency changes constantly,
providing linguistic information, as in the different intonation patterns
associated with questions and statements, and information about emotional
content, such as differences in speaker mood. In addition, the fundamental
frequency pattern determines naturalness of utterance production. This can be
illustrated by creating a synthetic version of a natural utterance in which the
spectral properties are left largely unchanged while the normally varying
fundamental is replaced with a fundamental of constant frequency.
The supralaryngeal vocal tract, consisting of both the oral and nasal airways
(Figure 1), can serve as a time-varying acoustic filter that suppresses the
passage of sound energy at certain frequencies while allowing its passage at
other frequencies. Formants are those frequencies at which local energy
maxima are sustained by the supralaryngeal vocal tract and are determined, in
part, by the overall shape, length and volume of the vocal tract. The detailed
shape of the filter (transfer) function is determined by the entire vocal tract
serving as an acoustically resonant system combined with losses including
those due to radiation at the lips. An idealized filter function for the neutral
vowel // is shown in the center panels of Figure 2 for a supralaryngeal vocal
tract approximately 17cm long, approximated by a uniform tube. The formant
frequencies, corresponding tothe peaks in the function, represent the center
points of the main bands of energy that are passed by a particular shape of the
vocal tract. In this idealized case they are 500, 1500 and 2500 Hz with
bandwidths of 60 to 100 Hz, and are the same regardless of the fundamental
frequency (i.e., they are the same in both the top and bottom center panels).

Figure 2: The source-filter model of speech production.
The source spectrum represents the spectrum of typical glottal air flow with
a fundamental frequency of 100 Hz. The filter, or transfer, function is for an
idealized neutral vowel //, with formant frequencies at approximately 500
Hz, 1500 Hz and 2500 Hz. The output energy spectrum shows the spectrum
that would result if the filter function shown here was excited by the source
spectrum shown at the left.
The spectrum of the glottal air flow, which has energy at the fundamental
frequency (100 Hz) and at the harmonics (200 Hz, 300 Hz, etc.), is plotted at
the top left of Figure 2. The amplitude of the harmonics, which for the
purposes of this figure combines the effects of both the source spectrum and
radiation, decreases by approximately 6dB per octave. At the top right of the
figure is shown the spectrum that results from filtering the laryngeal source
spectrum at the top left with the idealized filter function shown in the center
of the figure. Note that the laryngeal source has been "shaped" by the filter
function. Energy is present at all harmonics of the fundamental frequency of
the glottal source, but the amplitudes of individual harmonics are determined
by both the source amplitudes and the filter function. The bottom half of
Figure 2 shows the effect of using a different source function, while retaining
the same filter function. In this case, the fundamental frequency of the glottal
source is 200 Hz, with harmonics at integer multiples of the fundamental (400
Hz, 600 Hz, etc.). The spectrum that results from combining this glottal
source with the filter function for an idealized // has the same overall pattern
as that shown above it. However, there are differences in the details. Note, for
example, that the lowest formant for // has a center frequency of 500 Hz. A
glottal source with a fundamental of 100 Hz will have a harmonic at this
frequency. A source with a fundamental of 200 Hz will have harmonics that
straddle the lowest formant (i.e., at 400 and 600 Hz), as shown at the bottom
right of Figure 2. Since the overall shapes are the same, these details do not
change the perceived vowel quality, which would be that of an //. However,
the top example would be perceived to have lower pitch because of its lower
fundamental frequency.
The flexibility of the human vocal tract, in which the articulators can easily
adjust to form a variety of shapes, results in the potential to produce a wide
range of sounds. For example, the particular vowel quality of a sound is
determined mainly by the shape of the supralaryngeal vocal tract, and is
reflected in the filter function.Figure 3 illustrates this. Detailed accounts of
the acoustic properties of the vocal tract can be found in a number of sources,
including Fant (1960), Flanagan (1965), Fry (1979) and Lieberman &
Blumstein (1988).


Introduction | Acoustic Theory | Measuring Production
Tract Model | Gestural Modeling | State of the Art