Lecture 6-Clipping

Question 1

What’s the dynamic range of the human auditory system?

Answer 1

0-120dB

Question 2

Each bit we tag on to a digital audio system adds… dB of dynamic range.

Answer 2

6dB

Question 3

How much dynamic range do we get with 16-bits?

Answer 3

96dB

Question 4

Is 96dB of dynamic range sufficient?

Answer 4

For most applications, such as domestic listening is plenty.

Question 5

What ambient noise floor does a typical domestic living room have?

Answer 5

40dB
If we try to utilise the full dynamic range of an audio CD in such conditions: 40+96=136dB (beyond the threshold of pain!)
Typically having around 60dB of dynamic range is more than enough.

Question 6

What are the two main instances that the importance of large bit depth comes into play?

Answer 6

Listening to the final mix in excellent acoustic conditions (even here 16-bit is good), performing non-linear audio processing, such as compression and limiting. In these circumstances we a need a large dynamic range=more bits.

Question 7

If a sound is too loud what will occur?

Answer 7

Clipping

Question 8

What does clipping happen?

Answer 8

The audio equipment can only handle a certain input voltage range.

Question 9

Microphones change sound pressure into what?

Answer 9

Electrical signals.

Question 10

If we were using a 16-bit ADC we have 65536 (2^16) different amplitude levels between +/- 5V.

Answer 10

The maximum input signal, i.e. A signal hitting +5V is represented by the 65536th amplitude level, which we refer to as 0dB the max value that we can have.

Question 11

The reason it is 0dB is because our reference is 65536

Answer 11

So if we are at the max quantisation level of 65536, the output it:
Output level (dB)=20log10(65536/65536)=20log10(1)=0dB
This also works for the input voltage levels. Also assume our ADC only allows in signals in to 5V in amplitude. Then 5V is out ref for the electrical signals coming in. if the output of the mics hits 5V output level (dB)=20log10(signal voltage level/5)=20log10(5/5)=0dB

Question 12

What happens if we go about the 2^16 quantisation level?

Answer 12

Clipping, digital audio cannot go above the maximum level. Sine waves will start to look like square waves. Distortion will be introduced.

Question 13

In digital audio systems distortion is a very bad thing:

Answer 13

Results will be a lot of upper harmonics being generated that will make the sound nasty. Also harmonic distortion can also have other causes than just clipping.

Question 14

All audio circuits create a very small amount of distortion i.e. They ever so slightly change the shape of the original input waveform.

Answer 14

We can measure distortion by putting in a pure tone into an audio system and seeing how well it reproduces the tone at the output. We then look at the spectrum of the output and see if there has been any nasty upper harmonics created. If so, the system has distorted the input wave.

Question 15

Not all distortion sounds bad!

Answer 15

Analog values and tapes allow for soft-clipping, the distortion effects are not as harsh and are in fact psychoacoustically pleasant to listen to.

Question 16

Advantages of digital:

Answer 16

Wider dynamic range, best analog systems 80dB range, 16 bit digital:96dB, 24 bit digital: 144dB.
Increased resistance to noise: analog systems susceptible to crackling noise and hum from electromagnetic frequencies (EMF), thermal noise from analog components generate hiss. Digital is immune to this type of noise.
Better copy ability: analog copies degrade by at least 3dB per copy. With digital audio, unlimited generations of perfect copies can be made.
Error correction: CDs and DATs have built in error correction, cross-interleaved reed-Solomon coding (CIRC), 25% of a CD is used for error correction.
Durability:with analog each play causes microscopic particles of polyvinylcloride (vinyl) or oxide (tape) to be scraped away. Non-linear editing.

Question 17

Disadvantages of digital:

Answer 17

Bit rates, how many bits are used each sound to represent the signal, digital audio is expressed in kbps.
Sampling rate x resolution x no. Of channels =bit rate

Question 18

Digital mixing consoles:

Answer 18

They sample multichannel audio, process it, mix it and route it out of the console, all digitally and in real time.

Question 19

What does sample rate determine?

Answer 19

The maximum frequency component that can be faithfully recorded and reproduced.

Question 20

What does bit depth determine?

Answer 20

It determines the dynamic range of the recording.

Question 21

What is the connection between sample rate/bit depth and file size?

Answer 21

The greater the sample rate and bit depth the greater the file size.

Question 22

When does clipping occur?

Answer 22

When we drive the input signals of our microphones about the input operating range of the equipment.

Question 23

What is the result of clipping?

Answer 23

Harmonic distortion.