-
Source Filter Theory
- Acoustic theory of speech production
- Speech is a result from a combination of sound sources and the filtering provided by the vocal tract.
- an example of a forced vibration
-
Sound sources
- voice
- turbulent vibration
- transient vibration
-
Sources of Energy for Sound Production
Respiratory System
-
Respiratory System
provides the driving forces for both periodic and aperiodic sound production
-
Breathing for life
- approximately 12 times per minute
- inspiration - active contraction of the diaphraigm and internal/external costal muscles
- expiration - due primarily to passive forces as the elastic lung tissue gets smaller
-
breathing for speech
- controlled expiration
- need for constant air pressure and flow
- combination of muscular and elastic forces
- timing of breaths influenced by linguitic factors
- depth of breaths related to intensity, deeper for louder speech
- valves in the upper vocal tract use air pressure and flow to create sound
-
the larynx
the first valve in the vocal tract. creates periodic and aperiodic sounds
-
myoelastic-aerodynamic theory of voice production
- vocal folds are adducted to midline, using the muscles of adduction
- tracheal air pressure rises to a level sufficient to overcome laryngeal resistance and blow the folds apart.
- the folds are pushed upwarndly and laterally, increasing the area of the glottis
- vocal folds become maximally parted, resulting in a reduction in tracheal air pressure allowing the folds to return to midline
- the process repeats itself
-
medial compression
- force with which vocal folds are held together
- creates resistance to the flow of air through the glottis (space btw the vocal folds)
-
resistance is determined by
- medial compression
- longitudinal compression - force with which the vocal folds are stretched (cricothyroid)
-
glottis
space btw the vocal folds
-
Bernoulli forces
suction perpendicular to air flow
-
Elastic forces
tissue properties of the vocal folds
-
rest position of vocal folds
vocal folds together
-
rate at which vocal folds open and close
-
glottis area function
- shows a graph of the change in glottal area as a function of time during voice production.
- fundamental frequency
- period of glottal wave
- harmonic structure
- shape is saw tooth like
- open quotient
- open time / (open time + close time)
-
if the open quotient is lower, this means that
the vocal folds are closed longer for louder sounds
-
the glottal area function wave resembles
- the saw tooth wave the most
- use even and odd multiples
-
As frequency increases the glottis
spends more time open (less time in closed or rest position)
-
on the glottal are range a 1 means
that the vocal folds never quite come together
-
the louder you talk..
- the harder your vocal folds come together
- which can cause swelling, polyps, etc.
-
fundamental frequency
mean
- most common measure of voice
- relation to age
- relation to gender
- relation to musical scale
-
phonational range and registers
- pulse - lowest fundamental frequencies (aka glottal fry)
- modal - range most common for speaking and singing.
- includes musical terms of chest and head voice
- loft (falsetto) - highest fundamental frequencies
- phonational range - the range that a person can produce - typically 2-3 octaves (12 semitones per octave)
-
fundamental frequencies changes during spontaneous speech over the lifespan
- start together
- then both drop - males drop more significantly
- toward end of life males increase some and females decrease some.
-
fundamental frequency alone is..
not necessarily a good indicator
-
Determining range of vocal folds
n = 39.86 x log(high f0 / low f0)
n = # semitones
-
measures
- RAP = cut off .68
- Shim = cut off 2.81
-
fundamental frequency variability
- long term
- standard deviation
- range (2-3 octaves)
- short term
- perturbation - jitter and shimmer
-
perturbation
- cycle to cycle variability
- changes from one cycle to another
-
jitter looks at variability in terms of
period
-
shimmer look at variability in terms of
amplitude
-
pulse range / vocal fry is
the lowest fundamental frequency
-
when someone has a voice disorder
jitter and shimmer are expected to be higher
-
voice quality
- importance of harmonic structure
- harmonics to noise ratio - quantitative measure
- narrow-band spectograms - qualitative measure
-
waveform
- time between corresponding points = period (T)
- f0 = 1/T
-
wide band spectogram
- same as waveform
- vertical striations represent glottal pulses
-
narrow-band spectogram
- using harmonics to measure f0
- since f0 * 10 = f10 / 10 = f0. measure f10 or the highest visible harmonic that you can accurately count and divide by 10 (or the number of the harmonic)
-
narrow-band spectrum
- same as narrow-band spectrogram.
- using harmonics to measure f0
-
automatic analysis
use of software program to automtically measure f0
|
|