Intensity and Loudness Flashcards
MAP is measured in:
Close Field
MAF is measured in:
Open Field
Why is there a difference of 6 dB from MAF and MAP?
Reasons for the differences could be due to:
Internal Noise?
External ear resonance?
Binaural summation?
We found that: If calibrated in real ear, no 6 dB diff
What is RETSPL?
- Is the Normal hearing threshold in SPL (against 20 uPa in air across frequency)
- HL: using RETSPLs as 0 dB (reference)
- RETSPLs and allowance for ambient noise level
How can we get the norm across different labs/clinics?
Set up Reference-equivalent threshold sound pressure levels—RETSPLs
* coupler calibration
* Round robin loudness balance (to estimate variation across individuals and earphones)
What is the reference for the intensity level? (dB IL)
1 pW/m2 =10-12W/m2
What is the threshold level of subjects with best hearing in Pascals?
dB SPL: 20 µPa
What is dB HL?
RETSPL, average of normal subjects
What is dB SL?
Individual thresholds
If 10 times change in duration resulted in 10 dB reduction in threshold, what if the duration change is 2 times?
If klog10 = 10 dB, then k = 10
10log2 = 3, because log 2 = 0.3
Total energy = power*duration
Power is intensity.
Decreasing power by half is compensated by doubling the duration in terms of total energy.
Decreasing power by 50% = 3 dB reduction in threshold
Large temporal summation in low frequency.
What are the impacts of signal duration on hearing? (5)
When a tone starting at 5 cycles was played:
*Tonality: started at 5 cycles
* Temporal integration improves threshold.
Suggests cochlea acts as an energy detector
* There is improved Discrimination
* Perceived change in Loudness
* Improved Acoustic Reflex
What is Weber’s law (fraction) in terms of intensity discrimination?
Weber’s law: deltaI/I = constant
I is intensity in physical unit
What do these graphs show?
Longer duration of tones (clics) will create narrower spectrum and better frequency selectivity
If we use simple tone:
Weber’s law is not followed which we call Near-Missing tone in intensity discrimination
What is Weber’s fraction if expressed in dB?
deltaL(dB) = 10log[(I +deltaI)/I]
If intensity discrimination is 10% in Webers fraction, what is the threshold in dB?
10log[(I+deltaI/I)/I] = 10log1.1= 0.414 dB
We need more than _______in Weber’s fraction to be able to discriminate frequency in intensity discrimination.
20% in Weber’s fraction to be able to discriminate frequency in intensity discrimination.
What do these three intensity discrimination tets show?
in B) our intensity discrimination skill relies upon short memory, the pause duration (interval) matter in Pulsed Tones
in C) Long lasting signal may cause adaptation with continuous signals (pedestal)
The Near missing of Weber’s law for pure tones is:
Similar across different frequencies
Weber’s law is well followed by:
Wideband signals
Why small Weber’s fraction at higher SL for pure tone and spread of excitation?
The spread of excitation explains the near-missing of Weber’s law for pure tones.
Dashed line: spectrum of notch noise. The notch is where the auditory channel is open for signal. All other regions are masked.
Solid line: the spectrum of the signal presented in the notch frequency
What are the effects of Duration on intensity discrimination?
deltaL decreases with T ( -0.25log T)
Hold from 250 to 8000 Hz; from 40 to 85 dB SPL
Long integration up to a duration of 1 s !
Explain the Intensity Discrimination with Complex Signals analysis.
In this study, the research used Standard presentation: 21 components, widely separate in frequency, equal in amplitude
Standard + Signal, one component is stronger than others
Subject must compare levels in at least two different regions
Better results than single tone discrimination
What is happening in this task of intensity discrimination?
One component is changed in one interval
Subject is asked to tell in which interval its not flat anymore (which interval is louder than others)
(b)-(f): Roving level presentation—the overall levels varied between two intervals.
This ensures comparison on profile, or intensity difference across frequencies, not overall level.
When profile analysis is used (results of squares):
- Better discrimination.
- No impact of ISI.
Results in profile comparison are better (smaller thresholds) and not changed with the time delay (between the intervals). Not decayed with short-term memory
solid circle: 1k tone; open circle: 21 tones change intensity together; squares: 21 components with amplitude change only at one of them. Open squares: random level at each presentation (the same between two intervals), solid squares: roving level across intervals.
How do we quantify loudness?
Phone scale and equal loudness contours
Narrower dr at lower frequencies (takes less range to go from lowest loudness level perceive to the terminal threshold, Wider in the middle frequencies
Steeper at higher and frequencies
When do we use dB A, B and C?
dB A 40 dB (in quiet, class) widely used
dB B, 70 dB
dB C, 100 dB (Airport)
What does this graph show?
How loudness increases with sound level
At 1 kHz, for tone >40 dB, the level increase required for doubling loudness is 10 dB
Power law is followed only when the sound level is well above threshold.
Loudness increases faster near the threshold
But at lower levels:
Level increment for doubling loudness at lower level is smaller by a factor of 2 ~ 3 dB/10 dB:
What is the Power Law?
Power law
L (sones) = kI0.3= kp0.6 for sound level above 40 dB SPL
What does this graph show?
Loudness grows faster at lower sound levels near the threshold
Loudness-Intensity relationship
Both loudness an intensity are in log
What does this graph show?
Effect of bandwidth on loudness
If we apply sound close to threshold, we will not see the effect of critical band
suprathreshold phenomenon/issues
As the bandwidth of noise increases at constant intensity, the 1 kHz tone must increase its level to be equally loud as the noise.
As the bandwidth of noise increases at constant intensity, loudness increases
But below CB, no such change
What does this graph show?
Effect of bandwidth on loudness
Loudness increase with F, tested in two subjects, with two types of signals compared to the standard (standard tone)
Below 200 Hz: the seperation between bandwith does not show an increase of noise
Over 200Hz: increase of noise/loudness
Larger bandwith and larger separations can be seen over the CB
What does this graph show?
Temporal integration of loudness
Improved loudness with longer duration of signal
SPL required for equal loudness decreases with duration
200 ms upper limit works
The temporal summation becomes smaller at higher sound level
What is the use of uniform excting noise (UEN)?
UEN: the intensity is the same in each critical band, but not in equal frequency range in linear scale
What is a critical band?
Band of audio frequencies within which a second tone will interfere with the perception of the first tone by auditory masking
What is difference between UEN and white noise?
CB increases with frequency: higher the frequency, wider the CB. This is why the total energy in each CB for white noise increases with CF.
Therefore, in order to have equal intensity in each CB, the density of sound in high-F CB must be lower.
CB increases with frequency:
higher the frequency, wider the CB. This is why the total energy in each CB for white noise increases with CF.
In order to have equal intensity in each CB:
The density of sound in high-F CB must be lower.
What does this graph show?
Compare loudness growth function between 1 kHz tone and UEN:
Broadband noise is louder than narrowband signals; each of the two comparisons has bias
What is adaptation?
Response reduction, not due to the reduction of sensitivity
What is Fatigue?
Reduction of response due to reduction in sensitivity
How do we check for loudness adaptation?
Homophonic loudness balance
Heterophonic loudness balance
What does this graph show?
Partial masked loudness
How loudness change in noise
Masking elevates threshold
Reduced loudness at threshold
Same ceilling but different floor
Catches up at higher level
What is loudness recruitment?
Faster loudness growth with intensity
Occurs with subject with SNHL and normal hearing
Similar to masking effect:
Reduced loudness at small Sensation Level, catches up at higher level
Consequence: steeper loudness growth function
What are different methods of evaluation loudness recruitement? (3)
- Difference limen: in normal subjects DL is larger at low Sensation L; in SNHLs, DL is equal or smaller at low SL.
- Alternate binaural loudness balance (ABLB) for unilateral hearing loss
Present tones alternately to each ear
Avoid adaptation - Discomfort level (gives you ceiling)
What do these graph show?
Clinical categories of loudness recruitement
Complete recruitment :Catch up to the ceiling seen in SNHL
Incomplete recruitment: elevated ceiling Mixed HL
No recuitement: can’t catch up to ceiling CHL
What does this graph show?
Hyper recruitement
What can we say about the Intensity range in which recruitment occurs?
Traditional view: rapid increase at low sensation level (SL)
New findings:
Loudness at threshold (SL~0) is not zero for SNHL
Rapid increase at low-moderate SL, where cochlear compression should work in normal ear
How loudness recruitment in SNHL different from the effect of masking?
Non-zero loudness at threshold, faster increase at moderate levels
No-zero loudness, lost of appreciation of sound softness.
Lost of appreciation for soft sounds with SNHL
What does this graph show?
Loudness recruitement Relative BM response
The solid line shows a schematic illustration of an input–output function on the basilar membrane vibration for a tone with frequency close to CF. The dashed line shows a linear input-output function. This is typical of what might be observed for a tone with frequency well below CF (off-frequency tone).
You will not see an amplification of BM at on-freq region (6kHz) from an off-freq sound (3kHz)
What do these graph show?
Cochlear compression and changes in SNHL
What does this graph show?
The increase in response varied with input level: demonstrating compression
Gain is reduced at higher sound level from the active components in the cochlea
A larger input level increase (deltaM) is required for the same amount of output change (deltaO): the compression (lower gain) at high level
How can we evaluate cochlear compression in psychoacoustics?
By forward masking
- Listener detect a brief signal after a masker of 100–300 ms duration.
- Signal and masker are separated in time, so that there is no non-linear interactions (such as seen in two-tone suppression).
- Compare the effect of a masker with a Fre (in-frequency) close to the signal and a masker with a Fre well below the signal (off-frequency).
- Compression exists only in masker with Fre close to the signal!
- The difference shows the cochlear gain.
How fast is the foward masking for evaluation of cochlear compression in psychoacoustics?
- Conditions: masker is short (<300 ms, interval is short
- Signal and masker are compressed by a similar amount, the growth of masking function has a slope of unity (filled symbols). The slope is»_space; 1 using 3 k masker, due to linear growth of vibration by this tone (out of CF region)
Wider at low sound level because of the cochlear gain
What does this graph show?
Impact of SNHL (right)
Masking threshold elevated Masking effect match eachother between on and off frequencies which shows a lack of active gain
What is the relationship between, SNHL and loss of cochlear compression?
Cochlear becomes linear when dead or without OHC
The cochlear output (the response of auditory nerve) is not inconsistent with the hypothesis of cochlear compression
Central contributions must be considered
Origin of Loudness Recruitment in addition to the loss of cochlear compression:
Both positive/negative evidence exists of R-I of AN steeper in SNHL
OHC lesions-threshold elevation->more ANFs work in higher level
However, the overall output of ANF to brain is reduced.
Spread of excitation? denied by exp with notch noise
Central disinhibition: Very likely
What do these graphs show?
Cochlear damage always reduces cochlear neural output
R-I functions of AN are not steeper in impaired ear
Threshold increases, max spike rate decreases
Enhanced central responses at low frequency region
Increased “central gain” have been widely reported in subjects with HL
What is the relationship between central inhibition after SNHL?
Loss of central inhibition after SNHL
*The increased central gain is resulted from the loss of inhibition
*Reduced inhibition demonstrated as:
Reduced GABAergic inhibition in presbycusis, and other hearing loss
Reduction in GABA neurons, GABA concentration, GABA-R, and change of enzymes related to GABA metabolism
