Quantisation

[Carol F]

In the Sampling and quantisation video (slide labeled Concepts: sampling and quantisation) it states “Vertical axis: Continuous value (representing sound pressure) is encoded as one of a fixed number of discrete levels“. I’m not sure what this means.

[Simon]

It’s because each sample is stored as a binary number with a fixed number of bits. Let’s use 4 bits, which would give only these possible numbers (with decimal equivalents):

0000 = 0
0001 = 1
0010 = 2
0011 = 3
0100 = 4
0101 = 5
0110 = 6
0111 = 7
1000 = 8
1000 = 9
1001 = 10
1010 = 11
1011 = 12
1100 = 13
1101 = 14
1110 = 15
1111 = 16

That means that each individual sample will be quantised into one of those 16 possible values (i.e., amplitudes). No “in-between” values are possible.

Using more bits means more values are possible. The standard value in consumer audio is 16 bits. In music production, 24 bits is common.

How many possible values are there with 16 bits? What about 24 bits?

[Author]

Posts