Frequency and Pitch Flashcards
What is the meaning of auditory channels and bandpass filters? (3)
- Filters arranged in parallel in the cochlea with varied CF
- Spectrum peaks in the signal are detected by certain filters in the bank
- The bandwidth of the filter determines the critical band and frequency selective
The critical bands are determined by ______________________________
the filter bandwidth of the afferent neuron
The auditory channels are established as _______________ and are the basis for _________________________________
place code and are the basis for frequency analysis in the auditory peripheral
What is a bandpass filter/auditory channel?
Each channel of ANF responds specifically respond better to a frequency of sound
What are the two methods for evaluating auditory channels and frequency discrimination?
pulsed signals and frequency modulation (FM)
Why can we not use a continuous pedestal for frequency analysis?
Cannot present two tones simultaneously for frequency discrimination, because of pitch fusion.
What is frequency splattering?
Abruptly starts or stops transmitting a tone, will emit noise at frequencies other than the frequency of the sine wave.
What does this graph show related to frequency splattering?
We can reduce frequency splattering by a (slow) ramp or by masking with notch noise
What is shown in this graph related to frequency discrimination limen (delta f)
Little effect of intensity on frequency discrimination at low frequencies
deltaf constant below 500 Hz at low frequencies
deltaf increase with frequency (harder to discriminate two high frequency sounds from each other)
What does this graph show related to Weber’s fraction (deltaf/f) obtained with FM signals?
A slight decrease in intensity at high frequency, much larger at low frequency
Weber’s law: correct above 500 Hz; f/f = 0.7% =0.007
What are the results from non-FM signals (gated pedestal of tones)?
Pulsed tone, or band noise are seen with different cutoff
deltaf and deltaf/f are smaller than FM by factor of 3
Weber’s fraction: 0.2% (versus 0.7% for FM)
Interval (pause) improves performance in certain range!
Level dependent at low SL
What can we see from this graph related to the impact of sound level ?
The performance is improved with increasing SL at a low range, with no more improvement well above the threshold.
Discrimination ability: 0.2% of Weber’s fraction by the gated pedestal
What is the impact of the interval between two pulses in intensity discrimination?
The interval between the two pulses impacts the performance: the larger the interval, the larger the discrimination threshold—decay of short memory
What is the impact of the interval between two pulses in frequency discrimination?
The increase of the interval improves the performance in a certain range which is likely due to the reduction of pitch fusion with increasing intervals and for a longer intervals, performance will go down
What does this graph show related to the comparison between pulsed signal and FM (2)?
- Poorer performance using FM in frequency < 2000 Hz
- Pulse signals give better performance for frequency discrimination by a factor of 3 at the middle frequencies
- Better performance using FM in frequency > 2000 Hz higher frequencies than pulsed
How do we measure the DLF?
Two-tone pulses presented in a sequence (two pairs), subject indicates which of the two successive was higher in frequency.
DLC:
DLC: difference limens change: subjects indicates which pair differed in frequency in two successive pairs of two tone pulses (2IFC).
What does this graph show related to the effect of the stimulus duration on frequency discrimination?
Weber’s fraction is reduced with increasing duration, thus better performance, so an increase of frequency discrimination with increased sound duration
What are 4 methods to measure/define CB?
Threshold change
Masking procedure
Loudness sensation
Acoustic Reflex
Explain the method to define CB by threshold measure/ change: (4)
- Keep the total energy/intensity of sound the same
- Increase the bandwidth of signal from 0 to broad (BW) within the critical band (CB),the density of the sound will decrease
- This change in density will not change the threshold, if BW is in CB
- When the bandwidth goes beyond the critical band, some of the energy will spread out and wasted because the leaking to other bands will be too low and the threshold will increase. (need to boost up the sound to be able to hear the tone)
In the CB threshold method, increasing the bandwidth (BW) within the critical band (CB) causes the density of the sound to __________________, however, this change in density will not ________________________, if _________________.The threshold will ________________if BW > CB.
In the CB threshold method, increasing the bandwidth (BW) within the critical band (CB) causes the density of the sound to decrease, however, this change in density will not change, if BW < CB. The threshold will increase if BW > CB.
The masking threshold goes up when ______________________________________________
bandwidth is wider than CB
Define CB by masking procedure: (3)
- Only the energy of a masker in CB around the probe tone contributes to masking
- Keep the total intensity of the masker the same while increasing bandwidth from way below the CB to above CB
What will happen to the masked threshold of a signal when the BW of the masker > CB if the level of the masker remains UNCHANGED?
The masked threshold will go down when BW >CB, if the masker level remains unchanged and goes up when within CB
If you wish to keep the masked threshold the same with the BW masker change, then________________?
Increase the masker level
Define CB by loudness sensation: (2)
Keep the total intensity of two tones or a band noise the same
Increase the frequency distance between two tones or BW of the noise
Within the CB, the loudness will be the same but beyond CB the sound becomes louder
When BW> CB, loudness will?
Loudness will go up when >CB, because it is a suprathreshold issue.
When BW<CB, loudness will?
Loudness will go down when within CB
Define CB by AR:
Similar results to signal loudness change, when the signal goes beyond CB more auditory channels will be activated so louder sounds and acoustic reflex is stronger and thresholds will go down
How is the cochlea divided into Critical bands? (Lowest, Range under 500Hz, and beyond 500 Hz what do we use?)
Lowest CB = 80 Hz
CB=100 Hz to CF up to 500 Hz
Beyond 500 Hz we use the ratio scale to determine the CB = 20%CF, or 1/3 octave
CB = 20% or 1/3 octave of CF, therefore, each CB contains 100 Just Detectable Difference Frequency if _______________________
the frequency is above 500 Hz (linear ratio)
Unit of the critical band is :
Bark (in memory of Barkhuizen)
Crititcal Band:
1st: from _____________Hz,
2nd from _____________ Hz
From 0-16000Hz we have _______ steps
From 0-16000Hz, we have ___abutting CBs
1st: from 0-100Hz,
2nd from 100-200 Hz
From 0-16000Hz we have 600 steps
From 0-16000Hz, we have 24 CBs
What does this graph show related to CB width?
The division of the 24 CB by frequency is narrow at lower CF and wider at higher CF. Above 500 Hz, CB width is 20% of CF
What does this graph show related to the frequency-dependent function?
Top line: Shows how the Cb changes with frequency
Middle Line: The frequency change for 0.2 mm length of BM from CB
In each CB, there are _____ just detectable steps of _________________________________
6.3x4 = 25.2, in each CB, there are 25 just detectable steps of frequency change
How many hair cells do we have in each steps in CB?
we have ~3600 IHCs in each ear.
We have 6 IHCs in each step of frequency discrimination in CB (less than 20)
What is the neurological basis for Frequency discrimination? (5)
Place coding
OHC
Temporal Coding
Efferent Control
Central Inhibition
Place coding starts from _____________________________________
hair cells from ANFs then auditory channels with frequency selectivity, inherited in CAS
OHCs (active amplification) increase ________________________________________
OHCs (active amplification) increases the frequency selectivity of ANFs.
Temporal coding enhances____________________________________________________
Temporal coding enhances frequency coding in cochlea.
Efferent control to the cochlea enhances_________________________ and
suppresses ________________________
Efferent control to the cochlea enhances frequency selectivity and suppresses the impact of background noise
Central inhibition enhances ____________________ by__________________________________________________
enhances frequency selectivity by “masking” the response at edges, enhancing contrast
Important concepts about pitch: (3)
Sound induces pitch only when it is heard
2 frequencies that are too close (but > JDD) may produce identical pitch (pitch fusion)
Resolved vs. unresolved
Harmonic vs non-harmonic: difference in generating a pitch
Pitch fusion is _______________ than our frequency discrimination threshold which is why ______________________
Pitch fusion is much larger than our frequency discrimination threshold which is why we need to avoid it in clinic
Resolved versus unresolved (residual pitch):
Resolved can produce differentiable vibration on basilar membrane
Ex: if two signals fall in two different CB, why are they separated from each other
Unresolved: This could mean that there is no corresponding vibration in the cochlea OR the vibrations produced by two signals cannot be differentiated. (too close to each other or within the same CB
Give an example of unresolved (residual pitch):
Sensing the fundamental pitch but there is no corresponding vibration in the cochlea with that pitch
What is the difference between entities and partials in speech perception?
Entities are the overall impression of the whole sound like music played in a concert we appreciate the overall pitch of the instruments
Partials is pitch represented by individual instruments and different frequency components
What is the difference between integration and segregation in Pitch perception?
Normally, we can get integration without training but to hear partials you need to have good training
We are better at ________ pitch than _______ pitch.
synthetic than analytic because we need more training to hear the partials
Pitch vs timbre:
Timbre is the quality of sound which is the combination of Frequency components and
Temporal change (dynamic aspect)
Measurement of pitch:
Mel (stevens): 1000 mels by 1000Hz tone at 40 dB SL
No linear relationship with frequency
Double or half mels =
2000 mels or 500 mels
Meaning of mel scale Zwislocki:
mels are related to BM vibration, frequency position map; 4.5 mels to one f ( 6 IHC)
Meaning of mel scale according to Stevens, Davis and Lurie:
1 mel = a shift of 12 neurons or 0.8 IHCs
100 mel = one bark = 1.3 mm = 150 IHCs
Recall the number on slide 25, 1 CB covers:
0.2x6.3 = 1.26 mm.
100 mels - 1.3 mm along BM = 1 bark
With increased intensity, low-frequency pitch _____________, ____________ with middle frequencies, and ________________ at high frequencies pitch.
Low-frequency pitch decreases
no effect middle frequencies
high frequency pitch increases
At High frequency, when you increase the sound level you feel that pitch goes up
In order to not change pitch when IL increase:
Signal Fre must be changed:
At low frequencies, when IL increases we must increase the frequency to have the same feeling of pitch
At high frequencies, when IL increases we must decrease the frequency to have the same feeling of pitch
From this graph, we can see that starting at 1000 Hz you need to ________________ to feel the pitch at 2000 mels and there is a ________________________ between pitch and mels.
starting at 1000 Hz you need to increase x3 to feel the pitch at 2000 mels. There is a non-linear relation between pitch and mels
From the effect of duration on the pitch, a very short duration (<3 ms) makes the _______________________ and you need a duration of _______________ to have a pitch sensation
tone sounds like clicks
> 3-4 ms or 6 cycles is required to have pitch sensation
For the effect of duration on pitch, >10 ms for f>1000 gives a _____________________________ and _____________________
clear tonal sensation and improved over duration
For the effect of duration on pitch, >250 ms gives a ____________________________
> 250 ms, stable pitch sensation
What is the effect of ramping on pitch perception?
The longer the rise/fall time, the less frequency splattering and poorer transient, so better tonal sensation with slow ramping, longer duration
Zwicker’s place theory for pitch/frequency sensation:
Δf results in ΔE at a place, Fre dis vs Int dis.
Problems of place coding for pitch according to Achouten (1943):
Can’t explain periodic residual pitch: 800, 1000, 1200 Hz, the common denominator of 200 Hz dominates the pitch.
Problems of place coding for pitch according to Seebeck (1843):
periodicity of signal makes a contribution to pitch sensation
What we know about periodic pitch is that: (5)
- There is no need for place code
- Missing fundamental: interaction across harmonics causing temporal fluctuation (periodically) - Residual pitch processed in cochlear place
- Noise masking of low-frequency region has no effect, so low-frequency region is not required
- Broadband noise produces pitch sensation when modulated to produce periodicity.
- Temporal pattern (or periodicity) produces pitch; carried on by neurons with CFs as carriers frequency.
In CF modulation the term of amplitude modulation of carrier frequency means that:
When a tone of high frequency is modulated by a low-frequency tone, the high-frequency tone is the carrier, and the low-frequency tone is the modulator
In pitch perception, the missing fundamental is well explained by the ___________________. The ____________ breaks the rule of common denominator
Missing fundamental is well explained by periodicity theory. 60-Hz shift breaks the rule of common denominator.
What does this graph show?
Potential contribution from fine structure of the repeated waveform
I2 and I3: non-equivalent: the cause of ambiguities
What causes and why can we hear beating between two frequencies?
F1=1000
F2=1003
Beating: Two tones very close to each other, or small difference between them in phase and out of phase periodically
1000 Hz with an amplitude change
Can’t diff between 1000 and 1003 Hz
Pitch sensation of 2-tone combination:
when the difference is small…
when the difference is widened…
when the difference is further widened…
When diff is small, beat is produced (smooth modulation: 400+410 Hz)
When diff is widened (but <CB), roughness rather than discernible beats (400+440 Hz)
Further widening (>CB): separated pitches (400 +600 Hz)
To understand periodicity, Periodicity must be seen in:
cochlea, not only in the stimulation—in order to cause phase locking in ANF
To understand periodicity, separation of harmonics must:
not be larger than the width of CB—(or should be unresolved)
So that the two harmonics can interact with each other, causing periodicity in the CB (unresolved).
If the separation is larger than CB, you will hear__________________
separated pitchs
Problems of periodicity theory according to Patterson:
Phase change leads temporal pattern change, but not pitch
Problems of periodicity theory according to Hall and Peters:
Missing fundamental can be heard by sequential presentation of three harmonics (each 40 ms, interval 10 ms) in noise, but pitches of each harmonics in quiet
Problems of periodicity theory according to Goldstein:
pitch can be established by presenting different harmonics dichotically
Non-harmonic sounds can still produce pitch (such as dual-tone multi-frequency signal for telephone pads)
What are 2 physical Factors that Influence Pitch Perception?
- Onset Time: two sets of partials that have different onsets, will be segregated as different partials
- Harmonic Partials—Principle of the dominant component example in speech the
3rd, 4th and 5th harmonic component are dominant (if > 10 dB SL)
Not fixed to harmonic number but to the frequency range in which the sound is well resolved in the cochlea
What are the effects of modulation and co-modulation on pitch perception? (4)
- Modulation of one harmonic component breaks down the entity
- Fused pitch when tones to each ear commonly modulated
- Coherent modulation of all harmonic partials, harmonic remain
- Number of Harmonics: adding a new component increases the sense of entity
How does the spectrum affect pitch perception?
Unresolved high-frequency components impact residual pitch, Unresolved: either higher than hearing range or in the same CB.
Spectrum envelope: smooth spectrum favors single entity
Cont. on influencing factors on the pitch:
Tone Duration
Sound Pressure Level
Relative Phases: Least Important
Spatial Origin
Context Effects - Stream segregation
Subjective Factors that influence pitch perception:
Musical Training
Selective Attention
What are the theories related to pitch detection (2):
Spectral theory—two stages (1) frequency analysis (2) pattern recognition (spectrum)
Temporal Theory
neither of those theories can account for all pitch perception
What is this?
Pattern perception model which summarizes the steps of speech perception
What are the determining factors of Timbre:
Spectral factor—steady-state feature or tone color
Dynamic characteristics (separating percussive from blown instrument)—the role of signal envelope (temporal pattern).
Pitch fusion is _______________ than our frequency discrimination threshold which is why ______________________
Pitch fusion is much larger than our frequency discrimination threshold which is why we need to avoid it in clinic
We are better at ________ pitch than _______ pitch.
synthetic than analytic because we need more training to hear the partials
