Summary Chapter 8 p. 105-114
Speech acoustics: physical nature of sounds that travel through air from speaker to hearer.
Sequence of such sound = speech signal, forms output of speaking process and input to listening
process > speech acoustics can yield evidence about production and perception of speech.
Sound waves
Speaker makes sequence of movements of speech organs in production of normal utterance >
create disturbances in surrounding air > can be detected by ears of listener. Sound waves:
disturbances in the air that are detectable by the human ear > consist backward and forward
movements of air particles > create alternations of positive and negative pressure > transmitted
through the air in ever-expanding sphere from mouth to speaker > weakening until they die.
Sound waves are transitory and invisible > inaccessible to inspection and analysis > make
permanent recording of any speech signal. Magnetic tape allowed convenient recording of vast
amounts of speech data > nowadays recording are made digitally > transform into visual display
> identify patterns and make measurements.
Waveforms
Sound wave: represented by graph that shows changes of amplitude over time > waveform. Zero
point on vertical axis represents ambient atmospheric pressure > above and below line
represent alternations of pressure. Sound pressure waveform is audio waveform or raw
waveform > sound pressure waveform most basic. Waveform of whole utterance is more
variegated (succession of sounds instead of one) and more compressed along horizontal
dimension.
Waveform types
Four basic types > differences are result of different ways in which source of energy for sound is
generated in speaker’s vocal tract.
Periodic: repeating cycle > all nasals, vowels and approximants are sonorant sounds and have
periodic waveforms > source is periodic vibration of vocal folds.
Random: no repeating cycle > aperiodic > fricative consonants > source is noisy turbulence
created by forcing stream of air through narrow gap between articulators > sibilant fricatives /s/
and /sh/ achieve greater amplitude. Aspiration phase of voiceless plosive created by noisy
turbulence glottis also has random waveform.
Quiescent: silent pause in conversation can convey much meaning > significant at microlevel in
phonetics. Closure phase of voiceless plosives [p], [t], [k], articulators close off airway > vocal
folds don’t vibrate > no sound issues forth > waveform is straight line. Most speakers produce
voiced stops without vibration during closure in initial position > /t/ and /d/ differ in property of
voice onset time.
Transient: second phase of plosive > air pressure built up behind articulatory closure is released
with explosion > sound has aperiodic waveform but decays very quickly. No separate symbol for
transient on normal phonetic transcription.
Speech acoustics: physical nature of sounds that travel through air from speaker to hearer.
Sequence of such sound = speech signal, forms output of speaking process and input to listening
process > speech acoustics can yield evidence about production and perception of speech.
Sound waves
Speaker makes sequence of movements of speech organs in production of normal utterance >
create disturbances in surrounding air > can be detected by ears of listener. Sound waves:
disturbances in the air that are detectable by the human ear > consist backward and forward
movements of air particles > create alternations of positive and negative pressure > transmitted
through the air in ever-expanding sphere from mouth to speaker > weakening until they die.
Sound waves are transitory and invisible > inaccessible to inspection and analysis > make
permanent recording of any speech signal. Magnetic tape allowed convenient recording of vast
amounts of speech data > nowadays recording are made digitally > transform into visual display
> identify patterns and make measurements.
Waveforms
Sound wave: represented by graph that shows changes of amplitude over time > waveform. Zero
point on vertical axis represents ambient atmospheric pressure > above and below line
represent alternations of pressure. Sound pressure waveform is audio waveform or raw
waveform > sound pressure waveform most basic. Waveform of whole utterance is more
variegated (succession of sounds instead of one) and more compressed along horizontal
dimension.
Waveform types
Four basic types > differences are result of different ways in which source of energy for sound is
generated in speaker’s vocal tract.
Periodic: repeating cycle > all nasals, vowels and approximants are sonorant sounds and have
periodic waveforms > source is periodic vibration of vocal folds.
Random: no repeating cycle > aperiodic > fricative consonants > source is noisy turbulence
created by forcing stream of air through narrow gap between articulators > sibilant fricatives /s/
and /sh/ achieve greater amplitude. Aspiration phase of voiceless plosive created by noisy
turbulence glottis also has random waveform.
Quiescent: silent pause in conversation can convey much meaning > significant at microlevel in
phonetics. Closure phase of voiceless plosives [p], [t], [k], articulators close off airway > vocal
folds don’t vibrate > no sound issues forth > waveform is straight line. Most speakers produce
voiced stops without vibration during closure in initial position > /t/ and /d/ differ in property of
voice onset time.
Transient: second phase of plosive > air pressure built up behind articulatory closure is released
with explosion > sound has aperiodic waveform but decays very quickly. No separate symbol for
transient on normal phonetic transcription.
