Amp 1 Final Exam Flashcards
Define front end distortion and its cause?
it is distortion when the signal exceeds DR (usually a user’s own voice or loud music causes this)
Describe why digital hearing aids have a lower microphone dynamic range:
the analog to digital converter only has a 16 bit, which means that (x6) the 96dB DR is smaller
Describe how front-end distortion is managed in digital hearing aids:
dynamic range shift (cuts off soft sounds, so the mic is less sensitive to them), or increase the bit resolution (18 bit has a 108dB DR)
Prepare to calculate Signal to Noise ratio and understand its meaning:
SNR = signal - noise
- positive number is signal is more than noise, negative number is noise is more than signal
- Meaning: tells us the strength of the desired signal arriving to the person, and if it is stronger or weaker than the background noise
Bi-directional polar plot pattern and its directivty index
Bi-directional: nulls at 90 degrees on both sides
***** Directivity Index is the amount of attenuation, in SNR (low DI = low null)
Cartoid polar plot pattern and its directivty index
null at 180 degrees (DI = 4.8 dB SNR)
Hyper-cartoid polar plot pattern and its directivty index
nulls behind person on both sides, but has a small opening at 180 degrees (DI = 6dB SNR)
Super-cartoid polar plot pattern and its directivty index
smaller nulls at 120 and 420 degree spots, but a larger opening at 180 degrees (DI = 5.7dB SNR)
Be prepared to explain a polar plot in terms of sound collection and nulls:
A polar plot shows mic sensitivity as a 2 dimensional image above a person’s head. It also shows the null space, which are areas of attenuation.
Explain raising or lowering a TK and what that does practically in the environment:
- Raising a TK = makes soft sounds more quiet (up TK, down soft sounds)
- Lowering a TK = makes soft sounds louder (down TK, up soft sounds)
Practical applications for this adjustment: we would want to lower the TK for a person that wants to hear those soft sounds if it helps enrich the environment. We would increase the TK for a person that wants to remove those “annoying” soft sounds if they perceive them as noise instead of a valuable signal.
Explain expansion and its purpose:
Lower the intensity of very soft sounds and environmental noise (cuts out the sound of the rustling leaves, refrigerator, and the turning of a page)
CR: less than 1:1
Describe the purpose of the test box, reference microphone, and coupler microphone:
The test box chamber eliminates reflections, keeps ambient noise low, and has a calibrated sound source and mic to keep everything standardized across all practices.
Reference mic: makes sure the intensity is standardized by calibrating the SPL output from the test box speaker
Coupler mic: listens to the sound coming from the HA
What is an HA-1 coupler used for ?
CUSTOMS 2cc coupler simulates the ear canal with custom products (use fun-tac to attach it)
What is a HA-2 coupler used for?
BTE also 2cc couples that has a tube that can attach an ear hook, and mimics the tubing that would go into the receiver
What are 0.4cc wideband couplers used for?
RIC, CIC, IIC (smaller cavity)
What is the Zwisklocki coupler used for?
research purposes only
What is the purpose of a standardized coupler?
All couplers are used as a safeguard to be able to repeat the measurements over and over without variability, not meant to mimic a human ear!
Why do we compare our device measurements against a spec sheet?
We compare a spec sheet to see if our in-office performance of the device is set to standards from the device’s manufacturer.
Explain the process of converting:
An analog electrical signal to a digital signal: splicing the analog signal up into a bunch of pieces and them quantifying them onto either a 1 or a 0, then mathematically manipulating them to match the analog signal in numbers
what is sampling rate?
moments in time of the analog signal, ex: the peak and trout of the sound wave
what is Nyquist frequency?
The sampling rate is 2x higher than the frequency response. ex: 20k hz sampling rate becomes a 10k hz bandwidth (frequency response)
to accurately reconstruct a continuous signal from samples the sampling rate must be twice the highest freq present. The nyquist frequency is the highest frequency
what is quantization?
mathmateically rounds each sample creating a digitized version of the signal. Results in a less idealized representation of the analog input. Characterized by Bit values. increase bit = increase resolution
what is a quantization error?
the noise floor, it decreases when we increase the bit
what is processing speed?
the time it takes for the acoustic signal to travel through ALL of the components and then into the ear canal, usually around 2-10 msec
how does a digital signal go back to electrical?
either straight to the receiver (best, cleanest, less noise) OR first through the AGC-O and then to the receiver if it’s a high output signal
Explain front end limitations associated with 16-bit processing :
How does this impact microphone sensitivities?
Mics can’t detect high sounds, causes distortion (peak clipping)
Explain front end limitations associated with 16-bit processing : How are these limitations resolved?
Shift the DR up to allow those higher sounds in without having to distort them, but this does cutoff lower input sounds
The function of auditory filters (critical bandwidths) in the cochlea?
we have 25 bands that allow us to pay attention to frequencies within each band and ignore other frequency signals
- low frequencies have more narrow bands (160Hz wide)
- high frequencies have more broad bands (2,500Hz wide)
The effect of low frequency masking on the damaged cochlea ?
upward spread of masking masks the low frequencies as well as the higher frequencies at that same intensity
what frequencies are impacted by noise?
below 1500Hz
Function/purpose of automatic microphone switching:
switch to directional when there is background noise + back to omnidirectional when it’s quiet
Function/purpose of adaptive directional mics:
automatically change the polar plot to put the null somewhere else if there is a speaker in that area (ex: null moves behind you in a car if the person next to you is speaking)
There are different nulls for different frequency bands
Practical programming decisions related to adaptive vs. fixed directional microphones
Adaptive are best for only a few noise sources in the null
Fixed is best for many noises in the null space, because the hearing aid will know to cut out all of it instead of changing the polar plot
Function/purpose of adaptive beamforming mics:
narrow the polar plot to focus more on one direction
Louder = narrower bandwidth
what is the only way to increase speech intelligibility?
THE ONLY WAY WE CAN INCREASE SPEECH INTELLIGIBILITY IN NOISE IS: DIRECTIONAL MICS (not digital noises reduction)
Differentiate modulation rate and depth for speech and noise.
Rate: the speed that a signal fluctuates (speech is slow, noise is fast)
Depth: amplitude changes in the signal (speech is varying, noise is steady)
How is poor SNR determined in a hearing aid?
The noise is louder than the speech, which drowns out the speech (has a slow modulation rate, high depth)
If there is noise in speech and the amplitudes are similar, then the SNR is low and that’s a difficult listening situation… however if the speech has more depth, then the SNR will be higher and we will get a good listening situation
Digital noise reduction: describe the attenuated signal
One with a low depth, and a fast rate (ex: fan/AC, vacuum, engine idle, hairdryer)
Describe the benefits and limitations of digital noise reduction
Benefits: listening comfort, directionality, lessens listening effort and cognitive load
Limitations: takes away some speech noises as well (if there are high frequencies in the speech and you are reducing noise, it can also cut off some of those high frequencies from speech)
The theory associated with low frequency output reduction:
“noise” is in the mid and low frequencies, so if we remove those low frequencies, we could hear the higher frequency speech sounds better…
Limitations of low frequency output reduction:
noise also exists in the high frequencies too, if your undesired signal is also speech, taking out those low frequencies will also take out the desired signal low frequencies too
Explain Weiner filter function and limitations:
a spectral subtraction method (subtracting the noise from the speech)
limitation: takes too many critical speech sound out of the signal