Hearing Aids and Implants l Flashcards
Fill in the blanks related to the types of Hearing Loss:
speech information is carried by _____________________________________ (reflecting the movement of the articulators)
speech information is carried by the general shape of speech sounds (reflecting the movement of the articulators)
Each phone requires
__________________________
__________________________
intelligibility is best predicted by ______________________________________
Each phone requires
4 audible octaves or
4 speech features
Intelligibility is best predicted by audibility across a wide range
Hearing aids provide _______________________________________________
audibility across a wide range of frequencies
What does this graph show?
Threshold Shift: Loss of Audibility
We are asking ourselves is this person will be able to understand speech, sharp tunning curve etc.
Looking Closer (SNHL): Audibility and Tuning
Loss of tuning in SNHL
We can’t deal with the threshold shift unfortunately
What should we take into consideration when fitting HA?
Loudness Discomfort Level
You must take into consideration this level when fitting HA, so we need to estimate or predict it
Gain is:
how much things are ‘turned up’
how much sounds are turned up
How much gain to hear 0 dB at 1000 Hz?
Need 30 dB gain (which is a lot of gain to give someone to reach a 0 threshold)
How much gain to hear 0 dB at 4000 Hz?
What does this graph show?
What happens when someone speaks loudly (average level about 75 dB)?
What about the sound of the HVAC?
If the gain was already at 40 dB to reach 0 dB threshold, and someone speaks loudly, you’ll feel 115 dB level which would be TOO LOUD thus far too much gain.
Thus, we don’t really boost to 0 dB.
Speech is about ____ dB at 500 Hz and ____dB at 4 kHz
speech is about 65 dB at 500 Hz, 45 dB at 4 kHz
Speech has a dynamic range of about ____dB
30 dB
Should We Turn All Frequencies Up?
No, because boosting a certain range could cause other frequencies uncomfortable
What does it mean to use Frequency-Dependent Gain?
HA will have this option where you can shape the gain across the frequencies to fit speech in that range
What is compression?
different gain for different input levels
More gain for softer sounds
Less gain for louder sounds
What does this graph show?
It shows how much gain we predict to add speech by taking into consideration discomfort level
What are three features HA have to help shape speech safely?
Frequency-dependent gain
Compression
Output limiting
What are two fitting targets?
DSL
NAL
What is the DSL Approach?
How does the NAL work?
Can even calculate the SII
What is Compression specifically and what could it mean?
range of input levels is compressed into a smaller range of output levels
this could mean:
more gain for soft sounds (better for hearing pins drop)
less gain for loud sounds (more comfort)
What is Compression specifically and what could it mean?
range of input levels is compressed into a smaller range of output levels
this could mean:
more gain for soft sounds (better for hearing pins drop)
less gain for loud sounds (more comfort)
How does compression work?
when a loud input comes in, compression reduces the gain
so compression tends to make things softer
compression + gain increase can make soft sounds louder
What are the two types of compression?
Slow and Fast compression
Describe slow compression:
automatic volume control
can make sounds softer in loud environments
It doesn’t affect the spectral shape
gain is reduced by compressor, compression never has time to release (at least not until speech is finished)
Describe fast compression:
changes dynamic range of speech
can make the softer parts of speech louder
Compression is done in ____________________
Frequency channels
Give an example of how compression is done in Frequency Channels.
Consider the word ‘shoe’
Compressor would turn down gain after intense /∫/
Changes the dynamic range of speech
How does slow compression Affect the spectral shape?
Slow compression doesn’t affect spectral shape
gain is reduced by compressor, compression never has time to release (at least not until speech is finished)