exam 3 Flashcards
(145 cards)
1
Q
Describe four key points about perception
A
1: perception is deceptively hard
2: perceptual problems are usually ill-imposed (meaning you do not have enough information)
3: perceptual systems must be invariant
4: perception is unconscious inference
2
Q
Define psychoacoustics
A
determines the relationship between listeners’ behavioral responses to sound and the physical properties of sound
helps us relate the physical properties of sound to our perceptual experience, and can also help us understand the constraints of the physical auditory system itself
3
Q
psychometric functions
A
asks a listener to respond “YES” if change is detected, as a function of a stimulus parameter
“threshold” in this case is the stimulus parameter producing a criterion proportion of “YES” responses
4
Q
absolute threshold
A
the smallest value of a stimulus parameter that a listener can detect
a plot of just barely audible tones of varying frequencies
5
Q
examples of ill-imposed problems in perception
A
cocktail party problem
presence of other speakers obscure structure of target utterance, but speech remains intelligible
present-day speech recognition algorithms fall part in such circumstances (like your phone)
6
Q
MAP in audibility threshold
A
MAP: Minimum audible pressure
measured with inserts at the ear drum
7
Q
MAF
A
minimum audible field
measured with over-ear headphones or speakers
8
Q
humans can hear from – to ——- Hz but sensitivity diminishes below — Hz and above —- Hz due to impedance matching in middle ear
A
20 to 20,000
500, 8-10
9
Q
perceptual dynamic range of human hearing is ~
A
120 dB
10
Q
temporal integration
A
the process by which a sound at a constant level is perceived as being louder when it is of longer duration
suggests energy is added across time to detect sounds more effectively
11
Q
describe the relationship between power, energy, and duration
A
power is energy per unit time: P = E / T
12
Q
what is weber’s law
A
just noticeable difference (JND) between two stimulus is proportional to the smaller value
the larger the initial stimulus, the larger the change required for a difference to be noticed
13
Q
human discrimination perception uses ——— not absolute changes
A
RELATIVE
14
Q
if duration is ——- (and power stays the same), energy is ——— and thus goes up by – dB
A
Doubled, doubled
3 dB
15
Q
psychophysical tuning curves
A
frequency map of the masker sound levels needed to mask a fixed signal
16
Q
masking pattern
A
frequency map of the signal sound levels that are just detectable in the presence of a fixed masker
17
Q
upward spread of masking and physiological correlates
A
Explained by the asymmetry of the traveling-wave response pattern on the basilar membrane
18
Q
masking by noise can also characterize the ———- ——- used by the auditory system to process sound
A
bandpass filters
19
Q
signal-to-noise ration (SNR)
A
an approximately linear relationship between tone threshold and noise level (when masking of tones by broadband white noise)
20
Q
power spectrum model (theory) of masking suggests…
A
listeners pick the most advantageous filter to listen through for a given signal
21
Q
a tone masks tones of —— frequency more effectively than tones of lower frequency due to…
A
higher
upward spread of masking
22
Q
—– tones or ———- sounds close together in frequency mask each other —- than tones widely separated in frequency
A
pure
narrowband
more
23
Q
the greater the ———- of the masking tone, the —— the range of frequencies it can mask
A
intensity
broader
24
Q
each basilar membrane acts as a….
A
bandpass filter
25
broadband noise fills up the internal auditory filter so it's the ------ bandwidth (BWaf) that matters
filter
26
narrower noise (--- < ---) has all power within the auditory filter, so it's the ------ BW that matters
BWn < Bwaf
27
narrow noise means ---- making (easier to detect) as ------ bandwidth (BWn) gets smaller
less
noise
28
frequencies with passband of audtiory filter, or ------- ---- are critical for masking
critical band
29
two functional benefits of binaural masking (ie listening with two ears)
1) spatial release from masking
2) sound localization
30
listening in noise (with maskers) can be much easier if the source and maskers are separated in ---- even when the SNR is poor
space
(i.e. spatial release from masking)
31
monotic
stimuli present to only one ear
32
diotic
identical stimuli presented to both ears, or no interaural differences for the signal and the masker presented to each ear
33
dichotic
different stimuli presented to the two ears
"ch" dichotic is for the "change" between the two ears
34
three main sources of information for sound localization
interaural intensity differences
interaural time differences
spectral cues from pinna filtering
35
ITD
when the distance to each ear is the same, there are no differences in when a sound reaches both ears. when a sound is off to the listener's side, the time will differ.
a sound wave coming from the side will hit the nearer ear first, producing a time lag between the two ears
36
ILD
a sound wave coming from the side will hit the nearer ear first, producing a time lag between the two ears
37
neurons sensitive to interaural level differences are found in the..
lateral superior olive
38
superior olivary complex processes what information?
binaural processing of ITD (medial superior olive) and ILD (lateral superior olive)
39
duplex theory of sound localization
ITDs utilized at low frequencies
ILDs utilized at high frequencies
40
cone of confusion
binaural cues cannot tell you whether a sound is coming from above or below you
41
overcoming the cone of confusion
turning the head can disambiguate ILD/ITD similarity
42
to determine elevation, we use the pinna filtering properties
our pinnas are -----------, and the sound bounces off them as it enters our ears. this can give us clues about...
asymmetrical
how high or low a sound is in space
43
head-related transfer functions
used for spectral cues (from pinna)
the amount by which different frequencies are attenuated or amplified on their way to the eardrum
44
dB scale measures the intensity or pressure of a sound, provides an -------- physical measurement of sound levels and is useful for technical and regulatory purposes
objective
45
phons
subjective (perceived)
measures loudness levels based on the human ear's sensitivity to different frequencies
46
the phon scale is referenced to a 1000 Hz tone. if a sound is perceived to be as loud as a 60 dB sound at 1000 Hz, it is said to have a loudness of
60 phons
47
sone scale
subjective unit of perception of loudness
adjust the volume so it sounds twice as loud as the standard
48
pitch
the auditory attribute of sound according to which sounds can be ordered on a scale from low to high
49
pitch is a ----------- measure
subjective
50
loudness sone scale usefulness
useful in fields like acoustics and audio engineering, where understanding how humans perceive sound is crucial
51
pitch is useful for
melody of a song
gender and age of a talker
segregating multiple talkers
prosodic information (ex rising tones for questions)
word meaning in tonal languages
52
we can think of the cochlea as a set of -------- filters. the filters at ----- frequencies are narrows (let through fewer sounds) that those at ----- frequencies
bandpass
low
high
53
why is it harder to detect the difference between 2000
and 2100 Hz vs. 200 and 300 Hz, even though both tones are separated by 100 Hz?
because frequency is logarthmically spaced on the cochlea
the ration between 200 and 300 is different than the ratio between 2000 and 2100)
54
for tones, there must be between ------ cycles of the periodic sound to establish pitch
3 to 9
55
missing fundamental illusion
the pitch listeners hear corresponds to f0, even if it is missing
56
two general theories for pitch perception
temporal theory
spectral theory
57
temporal theory
pitch can be estimates from the temporal perioditicity of complex sounds
temporal periodicity is unaffected by missing fundamental, so this can work
58
spectral theory
pitch can be estimated from the frequency spacing of harmonics
finding the best f0 to match the harmonic spacing (any random set of harmonics can produce pitch)
59
timing cues to pitch
for low frequencies, auditory nerve spikes are ------------- to the stimulus
phase-locked
60
auditory nerve fibers are unable to phase lock to signal fluctuations faster than ~ ------- Hz
4000
61
individual harmonics are only "resolved" in the ---- of the cochlea, but it's OK because we still have
apex
temporal
62
resolved harmonics
apical (low freqency) filters have small bandwidths, they only pass individual harmonics (pure tones
63
resolved harmonics allow ------ theories to work well for low harmonic numbers with --- salient pitch percept
spectral
most (best)
64
unresolved harmonics
basal (medium-high frequency) filters have larger bandwidths, they pass several harmonics
65
unresolved harmonics allow ------- theories to work well for medium-high harmonic numbers, but --- salient pitch
temporal
less
66
two cues to the frequencies in sound
the place of excitation in the cochlea (which is limited by the resolvability of harmonics
the frequency of firing (which does seem to limit musical pitch perception to tones below 4000 Hz)
67
timbre
psychological sensation by which a listener can judge that two sounds with the same fundamental loudness and pitch are dissimilar
68
timbre is conveyed by ------- shape of harmonics and other frequencies
spectral
69
two factors that determine timbre (although still not fully understood)
relative amplitude of overtones
envelope of sound (the way intensity changes over time)
70
how does the auditory system solve this problem?
(four things)
uses knowledge of the world to best guess how many sources of sound there are, and what they are
1. grouping cues
2. attention
3. streaming
4. filling in
71
grouping cues
onset/offset
harmonicity
frequency modulation
repetition
spatial cues
72
common onset/offset
the brain interprets abrupt amplitude changes as the onset or offset of a distinact source
gradual amplitude changes are instead interpreted as a single source changing over time
73
harmonicity
the brain assumes that frequencies that are harmonically related belong to the ----- acoustic events
when one frequency is not harmonically related to others, it ---------- perceptually
same
segregates
74
repetition
sounds can be segregated and identified if they occur more than once across different mixtures, even when impossible to segregate in single mixtures
75
spatial cues
sound localization cues also contribute to...
the grouping and segmentation of sounds
76
spatial cues
a pure tone in one ear masked by noise in the same ear can paradoxically be ======== by the addition of noise to the other ear. only works if noise in two ears is the ------
"unmasked"
same
77
streaming
the process of grouping sound energy across time
78
describe classic streaming phenomenon:
galloping rhythm as single process or two separate streams
alternating high and low tones
79
if the difference in frequency is ----- -------- or the time between tones ---- ------- , ---- separate streams are heard
too great
too small
two
80
the continuity effect (filling in)
when noise occurs during a gap in a atone, we hear the tone continue through the noise
81
filling in only works if the brain if ... and ....
inferring that the tone's absence is best explained by masking
if the noise contains frequencies close to those of the tone, and is sufficiently high in level
82
------ content is filled in when it is plausibly masked by noise
speech
83
filling in, which we heard for speech and the applause texture in class, is a good example of:
unconscious inference
(fills in information to complete scenes)
84
music is.... (4)
- an acoustic phenomenon (sound)
- a perceptual and cognitive phenomenon (our brains must interpret that sound)
- a social and cultural phenomenon (a collective, arguable human specific behavior)
- an individual phenomenon (shaped by expertise and altered by deafness and disease)
85
two universal aspects of music
scales/melody (organized sets/sequences of pitches & harmony concurrent pitches)
rhythm
86
What are scales?
a set of musical notes organized in pitch
many musical systems build their melodies and harmonies around a single scale
87
melody
sequence of single notes that vary in pitch and duration
88
harmony
combinations of simultaneous notes
89
intervals
two stimultaneous notes
90
chords
three or more simultaneous notes
91
what makes some source combinations of tones more pleasant than others
preferences for intervals are learned rather than innate
some cultures do not have musical harmony at all
92
rhythm
a non-random repetetive pattern in time
93
sensorimotor synchronization (SMS) /entrainment
the temporal coordination of a rhythmic movement with an external rhythm
94
an important aspect of entrainment/SMS is that must....
anticipate the next beat
95
NMA-Negative Mean Asynchrony
on average, people tap before the beat
96
people can tap along (entrain) to rhythms with the inter-onset intervals between...
100 ms (fastest) and 2 seconds (slowest)
97
larynx
buzzy sound source
vocal cords open and close when air is blown through them
98
changeable resonates that filer the sound we produce in the larynx
pharynx (throat), mouth, lips, nose
99
factors affecting phoneme production
prosody (stress, intonation)
emotional state
different speakers (accents, gender, age, illness)
100
intonation
variations in pitch across a phrase
101
problem in phoneme recognition
variability across different exemplars of same phoneme/word
102
solutions to problems in phoneme recognition
acoustic features
categorical perception
knowledge of words
visual input
priming from context
103
assessment techniques
behavioral
physiologic
104
behavioral assessment
patient has to respond (raise hand when you hear beep, repeats same utterances)
involves entire auditory system (including brain)
105
physiologic assessment
no response required (sit/lie, relax or sleep in armchair)
measures specific aspects of auditory function
106
hearing screening
quickly separates people who may have a problem (pass or refer)
refer them for comprehensive assessment
107
diagnostic assessment
comprehensive
determines FOR sure whether there is a problem and its exact nature
108
human auditory system can temporally integrate (add up) energy, but only up to...
~300 msec
109
UNHS
universal newborn hearing screening
- all infants at birth
- phsyiologic measures
110
goals of audiologic assessment
degree of hearing loss
type of hearing loss
configuration of hearing loss
impact of hearing loss on the individual
patient's needs
111
basic test battery includes:
case history
otoscopy
pure tone audiometry
speech audiometry
112
otoscopy examine the:
external ear canal for cerumen, growths
tympanic membrane for rubes, perforation, etc.
113
pure tone audiometry
determines the degress, type, and configuration of the hearing loss
uses "pure tones" (sinusoids) generated by an audiometer
114
air conduction testing
uses earphones
tests entire auditory system
115
bone conduction testing
uses bone oscillator or vibrator
tests the inner ear (cochlea) and beyond directly by vibrating skull bones and contents
116
conductive hearing loss
outer/middle ear damage
(outer hair cells)
117
causes of conductive hearing loss
cerumen occlusion
external otitis
TM perforations
ossicular damage
otitis media (ear infection, fluid in the ear)
middle ear tumors
118
sensorineural hearing loss
inner ear damage (inner hair cells)
usually permanent and cannot be treated
119
communication difficulties with sensorineural hearing loss
- problem with delcate cochlear mechanism (frequency and intensity coding)
- leads to loss of volume
- decreased speech understanding
120
caues of SNHL
age
noise exposure
trauma
genetics
maternal infections
structural malformations
illness/infections
ototoxic drugs
tumors of the VIII nerve
121
central auditory processing disorder
difficulty in challenging listening environments
poor performance on degraded speech tests
normal outer/middle/inner ear, normal AC and BC thresholds
122
three ways to repair conductive hearing loss
tympanoplasty
stapedectomy
tympanostomy
123
tympanoplasty
repair perforated eardrum
124
stapedectomy
repair broken middle ear bone or repair otosclerosis
125
tympanostomy
ear tubes drain middle ear in case of otitis media (common in kids)
126
Bone Anchored Hearing Aids (BAHAs
surgically implanted into the temporal bone, and
use bone conduction to send vibrations to the
cochlea
127
in cochlear implants:
An array of electrodes is inserted directly into the cochlea. When the electrodes pulse, they directly activate ---------- ------ cells.
An external processor worn over the ear connects via a ------- to the internal electrode array and controls the current
auditory nerve
magnet
128
cochlear implants are only used when...
hair cells are damanged or gone
129
as of last year, over --------- people worldwide now have cochlear implants
1 million
130
tone-on-tone masking can be used to map out....
psychophysical tuning curves
131
what is the significance of PTC being similar to neural tuning curves?
Suggests that perceptual masking is constrained significantly by the physiological tuning that begins in the cochlea, and is also present in neural responses
132
more masking when masker is...
below signal (M
133
less masking when masker is....
above signal (M>S)
134
white noise has a ---------- and ----- average power spectrum over its bandwidth
continuous and flat
135
-------- or auditory filter bandwidth can be estimated in what ways from tone detection in noise?
critical ratio
critical band
notched-noise, equivalent rectangular bandwidth (ERB)
136
most reliable was to estimate auditory filter bandwidth
notched-noise estimates
137
filter BW increases as ------ --------- inreases, consistent with ------- tuning BWs
center frequency
neural
138
filter BW -------- as sound level increases, related to the reduction of -------- as SPL --------- (protecting the ear
increases
outer-hair-cell amplification
increases
139
filter BWs are ------- in listeners with sensorineural hearing loss, consistent with ---------
broader
outer-hair cell damage
140
spatial cues for understnading speech in noise
head shadow
binaural unmasking
141
head shadow
better SNR in one ear
142
binaural unmasking
added advantage from neural processing and stream segregation
143
two equally loud tones can have ---------- intensity
Very different
144
two equal intensity tones can have ---------- loudness
very different
145
vowels are distinguished by their -------- or filter resonances
formants
