SPPA 6010 Review Guide

SPPA 6010 Advanced Speech Science

SPPA 6010 Review Guide Test 2

The Speech Chain

Outline the general sequence of events from formulation to perception

Physical Quantities

Basic vs. derived quantities
Scalar vs. vector quantities
Key derived quantities and their basic relationship
displacement, velocity, acceleration, force, pressure, intensity

Physical Acoustics

Introduction to Sound
Sound defined
Spring-mass analogy – elasticity and inertia
Model of air molecule vibration
Characteristics of simple periodic motion
Amplitude
Period/Frequency
Phase
Representing frequency
Hertz
Octave/semitones
Representing amplitude
Instantaneous amplitude
Peak amplitude
Peak-to-peak amplitude
Wavelength
Simple periodic vs. complex periodic vs. aperiodic sounds
Fundamental frequency/period
Harmonics, partials, overtones
Graphical representation
Graphic representation of sound
Time-domain graphs
The waveform
Waveform envelope
Frequency domain graphs
The spectrum
Amplitude spectrum
Phase spectrum
The spectrum envelope
Short vs. long term average spectra
The Spectrogram
Relation to other plots
Narrowband vs. wideband
The acoustic filter
High-pass vs. low-pass vs. bandpass filters
Frequency response curve/transfer function
Cutoff & center frequency, roll off rate, gain, bandwidth
Input spectrum vs. frequency response curve vs. output spectrum
Frequency response curve vs. amplitude spectrum
Resonance
Free vibration
Forced vibration
Resonance curve/frequency response curve/transfer function
Representing amplitude-revisited
Root mean squared amplitude
The decibel
Intensity vs. pressure
Relative vs. absolute
Additive vs. multiplicative
IL vs. SPL
Conversion of pressure units to decibels

Digital Signal Processing

Analog vs. digital

Sampling frequency
Sampling theorem
Nyquist frequency
Potential problems
Quantization
Bit rate
Potential problems
Decision making for digital sound recording
Recording devices
Microphones
File formats

Introduction to Source Filter Theory

Assumptions of source filter theory
Tube and valve analogy
Graphic representation
Source characteristics
Filter characteristics
Output characteristics

The Source: Phonation

Basic laryngeal anatomy
Key structures
Skeletal structure
Intrinsic musculature
Innervation of intrinsic laryngeal muscles
Vocal fold structure
Vocal fold adjustments
Position
Tension/length
Myoelastic aerodynamic theory of phonation
Theory assumptions
2-mass model
Aerodynamic characterization of phonatory behavior
Volume velocity
Laryngeal airway resistance
Phonation threshold pressure
Driving Pressure
Periodic and Aperiodic Components of Phonation
Phonatory parameters: Acoustic measures
Fundamental frequency (F0)
F0 control
Mean, variability
Various nethods for extracting F0
Measuring signal amplitude
Spectral Tilt/degree of harmonic roll-off
Acoustic/physical bases of voice quality
Breathiness
Harmonics to noise ratio
Other methods of periodicity: autocorrelation, peak to average ratio, cepstral analysis etc.
Spectral tilt
Hillenbrand et al. paper
Pressed voice & vocal effort
Roughness
Jitter
Shimmer
Vocal Register
Observing Glottal Behavior
Laryngoscopic examination
Glottal area waveform
Photoglottogram
Flow glottogram
Inverse filtering
Electroglottogram

The Filter: Articulation and Vocal Tract Geometry

Basis Tube Acoustics
Area Function
Frequency response curve/transfer function
Formant calculation for uniform tube
Vowels
Articulatory Phonetic Description of Vowels
Formant relations for vowels (know typical F1-F2 values for corner vowels)
Stevens and House (SH) articulatory model of vowel production
Key assumptions and parameters in SH model
Acoustic consequences of changing model parameters
Vocal tract area function-acoustic relations
Role of cues to vowel perception
Formants
F0
Duration
Spectral change
Key plots
Wide band spectrograms for corner vowels
Spectrum envelopes for corner vowels
Frequency response curves for various vocal tract configurations
Various F1-F2 plots
Plots from Stevens and House paper

Segmentals: Acoustic and Physiological Phonetics

You must be able to draw and interpret time domain, frequency domain and spectrographic representations of all sound classes covered in the course.

Diphthongs
Key components of diphthongs
Articulatory and acoustic patterns of English diphthongs
Acoustic cues to diphthong perception
Glides and Liquids
General features of glides/liquids
Articulatory features of /w/ vs. /j/ vs. /l/ vs. /r/
Acoustic features of /w/ vs. /j/ vs. /l/ vs. /r/
Articulatory variation with /r/
Unique articulatory configuration of /l/ and its acoustic consequence
Typical formant values
Nasals
Unique articulatory configuration of nasals and the acoustic consequences
Formant vs. antiformant
Acoustic cues to place of articulation
Approaches to assessing nasal resonance

Plosives
Articulatory features of plosives
Aerodynamic features of plosives
Acoustic sequence during plosive production
Voiced vs. voiceless plosives
Source characteristics of plosives
Acoustic cues to voicing
Acoustic cues to place of articulation
Fricatives and Affricates
Laminar vs. Turbulent Flow
Source characteristics
Filter features for fricatives
Acoustic cues to place of articulation
Acoustic cues to voicing
Fricatives vs. plosives vs. affricates
Connected Speech Processes
Combining speech segments to produce utterances
Graphic representation of sound sequences
Coarticulation
Anticipatory vs. carryover coarticulation
Acoustics consequences of coarticulation
Suprasegmentals
Measuring speech quantity/duration/rate
Acoustic correlates of tone/intonation
Acoustic correlates of stress

General guidelines for Test Preparation

The purpose of this guide is to clarify your scope of responsibility as you prepare for upcoming tests.

1. Typical Test Format: Tests will typically include a range of multiple choice, short answer and longer essay-type questions. Term tests will be ninety minutes in length. The final exam is two hours.

2. Tests are cumulative: All exams are cumulative. That means that Tests 2 and 3 will cover material from earlier tests. Test 2 will have a percentage of material that was covered in Test 1 and Test 3 is a comprehensive exam that covers the whole semester.

2. Make sure you have carefully reviewed all reading assignments. Although material covered in lecture will be emphasized, all information in your assigned readings is fair game. If there are exceptions, I will note them in class.

3. Make sure you are very comfortable performing basic calculations. You should be able to do the calculations we cover in class in your sleep (e.g. frequency, wavelength, decibel conversion, resonant frequencies of a uniform tube etc). While, I provide some examples of problems you will face, I urge you to generate your own homework assignments and solve the problems.

4. Make sure you are very comfortable evaluating AND drawing the various graphs used in sound physics. Just as you should be able to do calculations in your sleep, the same goes for the range of graphs routinely used in speech science. I consider graphic interpretation to be a critical skill that generalizes well beyond speech science.

5. This is a graduate level course and therefore the test will reflect that fact. Commensurate with graduate education, expect that parts of this test will require you to apply concepts covered in more novel situations. So make sure you stretch yourself as you gain command of this material.

6. Use the Lab CD as a study tool. There are tons of speech and sound samples on the disk. Load up files and look at sounds using different graphical displays. For example, find vowels and measure their formant values. Compare the spectrograms and spectra of vowels/sounds. In other words, get command of this material any way you can. Just because you won’t be using the software on the exam, it doesn’t mean this material can’t be a great study tool.