Speech and Hearing Science Exam 2

Sound

A propagation of a pressure wave in space and time

Elasticity

Opposing displacement

Inertia

Opposing acceleration and deceleration

Pressure

Force exerted over a unit area

Compression

Areas of high density and pressure

Rarefaction

Areas of low density and pressure

Period

Time it takes to complete one full cycle of motion

Frequency

Cycles per second

Wavelength

Distance covered by a high-pressure region and succeeding low-pressure region

Sinewave frequency

The number of full cycles occurring in 1s intervals

Sinewave amplitude

Displacement of an air molecule from rest position

Phase

Position of the sinusoidal motion relative to reference position

Sine wave

- periodic

- single frequencies, simplest acoustic event

- building blocks of complex wave forms

Waveform x-axis

time

Waveform y-axis

amplitude

Spectrum x-axis

frequency

Spectrum y-axis

amplitude

Complex periodic waveform

Sum of individual sinusoids at the harmonic frequencies

Complex aperiodic events

no repetitive patten and no harmonically related frequency components

Resonance

Object vibrates with maximum energy at a particular frequency (range), natural frequency

Damping

Energy loss in vibratory systems

What factors cause damping

Friction, absorption, radiation, gravity

Friction

- Objects rubbing against objects or structures, losing energy in the form of heat

- air molecules rub against one another and the walls of the resonator, generating heat and expending energy

Absorption

- Vibrating object transfers energy to another structure

- air molecules transfer energy to the vocal tract, making that vibrate

Radiation

- sound energy escaping (radiating) from the tube and being lost

- escape of sound energy from mouth and nose

Gravity

- can exert a force on the object opposing the forces inherent for vibration

- E.g. pendulum motion

Bandwidth

Index of tuning, range of frequencies between 3-dB-down point on either side of peak energy

How do air particles move?

If a force is exerted on the molecule, it moves and then comes back via recoil force (stored energy from displacement) and inertia. Recoil and inertia make the particles vibrate, friction stops the motion

Displacement

How far something moves away from its equilibrium position

How are waveform and spectrum different?

Waveform shows an acoustic event in time domain, while spectrum shows an acoustic event in frequency domain

All Waveforms

t/f: Almost anything can vibrate?

True

Spring Mass Model

simplified physical system that represents the relationship between a mass and a spring to study motion

Spring mass model- Period of Vibration

Amount of time required to complete one full cycle of motion

Spring Mass Model- Frequency

number of cycles in one second

Spring Mass Model: Mass

Takes more time to move a cycle- increases the period/decreases the resonant frequency

Spring Mass Model: Stiffness

Stiffer objects require greater force to be displaced, stiffer coils have greater recoil forces and thus greater rates of movement. This decreases the period/increase the fundamental frequency

Spring Mass Model: Mass and Stiffness

- increases in stiffness, increase the resonant frequency

- increases in mass, decreases the resonant frequency

- Both can vary independently

Helmholtz Resonators

- acoustic equivalent to spring mass model

- neck is the mass, if the mass is low we'll have a higher frequency

- If mass increases the frequency decreases

- Compliance and stiffness are inverse

Standing wave

- Put a wave through a tube, it's reflected at multiple points and it looks like the wave isn't traveling

- Points where your wave is at atmospheric pressure is in the same location

Tube Resonators: with both end open

- we have a standing wave

- sending different frequencies through the tube, certain ones will produce strong pressure changes

Tube Resonators: with one end closed

- standing wave in the tube

- open end of the tube has atmospheric pressure

- closed end of the tube has max pressure the wave reaches

Vowel acoustics

Vowels vibrate like a tube closed on one end

Signal is periodic

Rate at which it repeats is the fundamental frequency

- adult women: 90-200 Hz

- Adult men: 115-125 Hz

- 5 year old children: 250-300 Hz

Characteristics of the signal

- Slope of the opening phase is shallower than that of the opening phase

- Signal shows times where the vocal folds are open (60%) and apart (40%)

- complex periodic wave form, not a single sinus sound

Glottal source spectrum

- lowest frequency: fundamental Frequency (FO), first harmonic

- other frequencies are whole-number multiple of FO

Voice is a complex periodic event

repeating, non-sinusoidal shape

Voice is quasi-periodic

very small variations in successive glottal cycles

Period (and FO) changes

- depend on age, sex, etc

- spectrums of higher FOs are more spaced out

Frequency domain: Impact of the shape

the steeper the closing slope, the less tilted the spectrum

Hyperfunctional voices

-very rapid closing phase

- less than normal tilt

- "pressed" voice

Hypofunctional voices

- slow closing phase

- more than normal tilt

- weak and breathy voice

Frequency domain: Open and closed phase

- 60% open, 40% closed

- speed of closing is related time one/closed

- less open time, less tilted spectrum

- more open time, more tilted spectrum

Filter

Vocal tract as a tube closed on one end

Vocal tract resonates as a tube closed on one end

- during vibration the vocal folds snap shut

- airflow is blocked, air above the vocal folds is compressed and starts a pressure wave

- no airflow at the VF, airflow at the lips: tube closed on one end

- vocal tract resonances are excited with every snap of the vocal folds

Area function of vocal tract

plot of cross-sectional areas as function of distance from glottis to lips

Vocal tract configuration

-includes constrictions

- a combination of "tubelettes" for which the width can be calculated

Vocal tract shaping the input signal

source + filter

Multiplication of input and filter equal

output of the vocal tract

Impact of the filter on the source

- harmonics are shaped by the form of the filter

- resonant frequencies amplify energy; in valleys the energy is little to not amplified

- still a systematic decrease of energy in harmonics

First three front (peaks) are most important for

vowels (frequency information is the most important)

Formant Bandwidth Friction

Generates heat

Formant Bandwidth Absorption

When a frequency is close to the resonant frequency of a surround structure (tongue, cheek), energy is absorbed, which increases damping (widens bandwidth)

Formant Bandwidth Radiation

loss of energy going from enclosed tube to open environment

Formant Bandwidths

Factors responsible for energy loss are active in the vocal tract

Narrow Bandwidths

vowel resonances are fairly "sharply tuned"

Broader Bandwidths

perceived as "muffled"

Source filter theory

Sound at the level of the glottis + configuration of the vocal tract = the sounds we hear

Constriction located at pressure maximum

- raises resonant frequency

- constriction increases stiffness of air molecules

- the greater the constriction, the greater the increase in resonant frequency

Constriction located at velocity maximum/atmospheric pressure

- lowers resonant frequency

- constriction increases acoustic mass (inertia) of air molecules

- the greater the constriction, the greater decrease in resonant frequency

Constrictions have different effects on each resonant frequency

1) any constriction affects all resonant frequencies

2) a given resonance may be affected by two simultaneous constrictions

Three-Parameter Model of Stevens and House

- tongue height

- tongue advancement

- lip rounding

Three-Parameter Model of Stevens and House: Tongue Height

- F1 varies inversely with tongue height. The higher the tongue, the lower F1

- Relative height of the tongue at the location of the major constriction

- more pronounced for front vowels compared to back vowels

Three-Parameter Model of Stevens and House: Tongue advancement

- F2 increases and F1 decreases with increasing tongue advancement

- position of major constriction of vowel along anterior-posterior dimension of the vocal tract

- F2 increases as the constriction moves to front

- F1 decreases as the constriction moves to front

Three-Parameter Model of Stevens and House: Lip Rounding

- increased lip rounding decreases all formant frequencies

- lips to teeth as a separate compartment with an opening area and a length, lengthens the vocal tract

- some vowels are produced with lip rounding

F3

- third formant is not easily related to changes in articulatory dimensions

- only small changes seen

- only applicable rule on F3 is lip rounding

Spectrogram

x-axis = time

y-axis = frequency

darkness/color= amplitude

Narrowband spectrogram

have a high resolution frequency analysis, and are detailed enough to see individual harmonics

Broadband spectrograms

have a coarse frequency analysis, but more detailed time analysis. The vertical lines represent increases in amplitude with vocal fold closure

High front unrounded vowel i

- has a characteristic high frequency resonance, the consequence of a small oral cavity

- tongue fills the oral cavity and the pharynx opens up

- large pharyngeal cavity resonates at low frequencies (F1); the small oral cavity resonated at high frequencies (F2)

Low back vowel a

- oral cavity is large and the pharyngeal cavity is small

- tongue is lowered, either passively through opening the jaw; or actively by depressing it, or both

- small pharyngeal cavity resonates at high frequencies (F1); the large oral cavity resonates at low frequencies (F2)

High Back rounded vowel- u

- tongue is lifted to the roof of the mouth, opening up the pharyngeal cavity and lengthening the oral cavity. The lips are rounded and protruded (or the tongue is retracted more)

Diphthongs

- change resonance characteristics during its production

- articulation is characterized by two articulatory positions (onglide & off glide) Essentially a diphthong is a combination of two vowel sounds

Vowels

- resonators can be described as a single tube

- have wavelengths longer than the cross-section of the vocal tract

- complex periodic sound

Nasals are produced with

- velum open

- complete oral closure

- when these happen at the same time: nasal murmur

Nasal murmur

- pharyngeal-oral tube is closed and constricted similar to i

- low 1st formant

- low energy because of damping

- have antiresonance

- low dark bar on the spectrogram

Nasal sound: m

- articulation at the lips

- velopharyngeal port is open

- acoustic properties: nasal murmur

Nasal sound: n

- articulation at the alveolar ridge

- velopharyngeal port is open

- acoustic properties: acoustic murmur

Nasal sound: ng

- articulation at the soft palate

- velopharyngeal port is open

- acoustic properties: nasal murmur

Nasalization

- pharyngeal and nasal cavities are coupled

- both tracts are open to the atmosphere

- sounds come out of both mouth and nose

- both cavities have resonant frequencies

- still important even though we don't use it in English

Lateral sounds

- in articulation of /l/ air passes through two parallel air passageways

- air is trapped behind the closure, antiresonance with the greatest effect on F3

liquid /l/

-tongue tip contacts alveolar ridge

- sides of the tongue come down: lateral emission of air

- Acoustic properties: F3 is level for /l/

Obstruent sounds

- sound source between resonating cavities, noise is generated at the constriction

- vibration of vocal folds is shaped by vocal tract

Airflow in tubes

- air speeds up at a constriction

- frication source, which is shaped by the vocal tract

Source spectrum for fricatives

- source is more or less the same for all fricatives

- voiced fricatives have an additional low-frequency energy due to vf vibration

Shaping of sound source for fricatives

- the narrow constriction divides the oral cavity in two

- front cavity acts as resonating cavity

- back cavity acts as a "closed" cavity: antiresonance

Fricatives: [f] [v]

- labiodental fricatives

- articulation is at low lip

- virtually no resonating cavity anterior to the constriction

- low intensity frication noise

Fricatives: voiced and voiceless th

- lingua-dental fricatives

- articulation is at the tongue tip

- virtually no resonating cavity anterior to the constriction

- low intensity frication noise

Fricatives: [s][z]

- alveolar fricatives

- articulation with the tongue creating a constriction at the alveolar ridge. Air flows through midline groove of tongue against teeth

- short anterior cavity emphasized high frequencies

- narrow, high frequency, high energy noise

Fricatives: [sh][ʒ]

- palatal fricatives

- articulation as the tongue forms grooves in alvelopalatal region, lips are often rounded

- long anterior cavity emphasizes Lowe frequencies