# Exam 3

p. 118-172

## Terms

undefined, object
copy deck
fourier analysis
a math technique that reduces any wave (oscillation), no matter how complex to a series of simple sine waves
d0
as increases, f1 frequency decreases and f2 frequency increases
speech waveform
requires resonance, spectrum envelope, frequency and amplitude (dB)
A/le
As lip rounding increases, both f1 and f2 decreases
Peter Jennings
low duty cycle, richer voice, more energy, extra amplitude gives richer voice
glottal spectra
spectrum of source frequencies: f0 + harmonics
Resonance of uniform tube
Fn= NC/4L Fn= nth resonance of frequency, c= velocity of sound. L=length of the tube (resonator). N= odd multiples (1,3,5,7,....)
parameters to formant resonances
1. d0= distance from glottis to highest point of tongue constriction. varies 4 to 13 cm....2. r0= cross sectional area between top of tongue and roof of mouth. varies 0.3 cm to 1.2 cm......3. A/le lip rounding index, varies 0.1 to 20 cm.
approximants
non-nasal semi-vowels; dependent upon speed of VT change for proper production
complex wave
contain multiple frequencies
central vowel
no obvious point of constriction or at hard palate
spectra characteristics
1. place of constriction, 2. shape of orifice at constriction, 3. pressure drop across the constriction
Helmholtz resonator analogy
high resonance = small volume; low resonance = large volume
aspiration
glottal fricative /h/; delay in VOT
noise
random variation in air pressure
affricate
transient + noise sound source (stop + a fricative); major VT constriction at posterior of alveolar ridge
Source filter theory
Vocal tract acts like a frequency selective filter; VT resonates the signal at the voice source (Vg) by allowing certain frequencies to pass through the filter with greater amplitudes relative to other frequencies
r0
(tongue lowers), f1 increases and f2 decreases
front vowel
point of VT constriction near alveolar ridge
semivowels
dependent on speed of VT change for proper production
suprasegmental
pitch, intonation, stress, duration, rhythm, juncture; independent of and of longer duration than segmentals
low vowel
low tongue position, no constriction (largest r0)
reynold's number
Re= u*h/v; u= a constant; h= width (cm) of passageway; v= particle velocity (cm/sec); reynold's number is reached with the combination of passage width and flow velocty interact to cause turbulence and eddy formation from previously laminar air flow.
Vowel Phonetic classification
1. major point of vocal tract constriction (front - back position of tongue), 2. degree of vocal tract constriction (amount of tongue elevation), 3. degree of lip rounding (continuum from rounded to spread), 4. tense vs lax (vague underlying cause; tension?)
voiceless affricate
as in 1st and last sound in church; produced as a combination of voiceless alveolar stop /t/ and voiceless lingua-palatal fricative /s/
mid vowel
moderate tongue elevation
noise
aspiration, friction, affrication, +/- sibilance
three sound sources
voicing, noise, transient
simple wave
contain only one frequency, sine waves, triangle waves, square waves, whistle
Voicing
stops, fricatives, nasals, approximants, laterals, vowels
MM
big duty cycle, psg doesn't build up as much so there is more open time
back vowel
major constriction occurs between tongue dorsum and velum
complex periodic
character vowels, consists of the repetition of the same basic waveform, cycle by cycle... each cycle is called a pitch period
high vowel
most constriction (smallest r0)
output spectra
spectrum of harmonics of speech output waveform
complex aperiodic
do not have periodic or repetitive pitch periods corresponding to voicing. Energy present at all frequencies.
voiced affricate
as in 1st and last sound in judge; produced as a combination of voiced alveolar stop /d/ and voiced lingua-palatal fricative /z/
spectral envelope
the line that connects the tops of all the harmonic components
transient
stop release
continuous spectra
sounds that have noise (fricatives- s, z, sh, v, f)
line spectrum
output speech sounds, line corresponds to harmonic, must be complex speech b/c of frequencies

41