Compression Flashcards

1
Q

What is the aim of compression?

A

To fit a wide dynamic range of speech sounds into the narrow dynamic range of a hearing impaired person

Intense sounds need less amplification than weak sounds

A compressor is an amplifier that automatically turns down its gain as the signal level rises above a predetermined point

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Feedback compression

A

AGCO

Automatic Gain Control Output

Makes decision to turn the amplification down by looking at its output

More commonly used - then you can do MPO

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Feedforward Compression

A

AGCI
Automatic Gain Control Input

Amplifier looks at input to decide whether to kick in compression

Sounds comes in to the amplifier at a moderate level, not loud enough to kick in the compression, may be amplified too loud for the client, whereas output limiting then you choose what the highest level is that it will present to the client

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is attack time?

A

The time taken for the compressor to react to an increase in signal level

It is usually define by how long it takes for the compressor to reduce the output signal by 10dB

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is release time?

A

The time taken for the compressor to react to a decrease in input level

How long does it take, post input change, for the compressor to increase the gain by 10dB

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

If you are talking about your stock standard 70year old mild-moderate presbycusis you will want ….. attack time and ….. release

A

Fast attack, slow release

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Problems associated with long vs. short attack and release times

A

Short release: may hear an audible increase in the sound volume (artefact)

long: no artefact as gain increase is more gradual. But soft inputs after a loud signal will receive insufficient gain and may thus be inaudible

Release times need to be medium-long to avoid artefact (always over 20ms, usually 100ms)

If attack time too slow will be uncomfortable

Attack times in HAs are commonly ~5ms

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is phonemic compression?

A

Quick release, Quick attack

Good: reduces dynamic range/makes soft consonants audible and loud ones comfortable. Reduces the spectral differences between phonemes

Bad: Alters loudness relationships between adjacent consonants possibly making them more difficult to identify. Also in quiet times between speech any soft noise present will be amplified

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What release/attack times are used in Automatic Volume Control?

A

Slow attack, slow release

Good: phonemic loudness relationships preserved

Bad: soft phonemes may be inaudible, loud noises may be uncomfortable

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Discuss adaptive release time

A

Depends on signal being amplified

For speech release time is slower, for brief intense sounds it will be fast

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Define the Compression Threshold

A

The point at which the output deviates by 2dB from the output that would have occurred had linear amplification continued to higher input levels

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Define Compression Ratio

A

The change in input level needed to produce 1 dB change in output level

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is Wide Dynamic Range Compression?

A

Compression applied over a relatively wide range of inputs

Usually best achieved through the use of input compression

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is compression limiting?

A

Compression applied for only high level inputs

Usually best achieved through the use of output compression

Typically used with high compression ratios (greater than 8:1)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Discuss the difference between High level, WDR, and Low level compression

A
  • The problem with the high level - we now have a situation where we have stopped the distinction between moderate and loud sounds - both sound loud
    • Wide DR - we’ve maintained the perceptual difference of intensity of sound (soft sounds sound soft, loud sounds sound loud etc.)
    • Low level - distortion is occurring for softer sounds (soft sounds heard as moderate)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

…. provides WDRC and ….. provides limiting.

A

Input provides WDRC

Output provides limiting

17
Q

What is curvilinear compression?

A

Ratio changes continuously as input level varies

input/output graph is a curve

Doesn’t seem to be superior or inferior to other types of compression but is quite common nowadays

18
Q

What are the two classifications of compression schemes across frequencies?

A

TILL - when the degree of compression is greater in the high frequency channels there will be greater high frequency emphasis at low input levels (Most common compression)

BILL - when the degree of compression is greater in the low frequency channels there will be a greater low-frequency emphasis at low input levels (BILL)

19
Q

What are the 8 rationales for using compression?

A

Avoid discomfort, distortion and damage

Reduce inter-syllabic intensity differences

Reduce long term level differences

Increase listening comfort

Normalise loudness

Maximise intelligibility

Reduce noise

Empirically determined

20
Q

What tactic is used to maximise speech intelligibility?

A

Multichannel compression used to achieve in each frequency region, the amount of audibility that maximises intelligibility

21
Q

NAL vs DSL

A

Both use WDRC, both have a low ratio just NAL uses a slightly higher CT

22
Q

Speech intelligibility vs reduce noise

A

Speech intelligibility - slightly less compression (slightly higher CT) than standard WRDC

Noise - more compression (lower CT and higher ratio) in the low freq. than the high

23
Q

What does ADRO stand for and what is it?

A

Adaptive Dynamic Range Optimisation

Output based amplification

- Gain is reduced so level does not exceed UCL
- Gain is reduced so 90% of speech doesn't exceed UCL 
- Gain is increased so lower levels of speech (lower 30%) are audible 
- Gain is never allowed to exceed maximum

For people with limited/reduced dynamic range

About 10% of the time too loud, 30% of the time inaudible