top of page
Writer's pictureErika Camilleri

Chapter 45 Sampling Techniques (Digitisation)

Updated: Aug 31, 2023

We know by now that computers work with 1s and 0s, a digital language, but the world around us is not digital. The music coming from a piano for example, is an analogue signal. In this chapter we will understand the difference between analogue and digital, and how we can translate from one to the other.

What is an Analogue signal?

Let's consider sounds. Real-world sounds are continuous and represented in a physical way. We call these analogue (or analog). When the sound from a piano or someone singing reaches your ear, your ear is receiving an analogue sound. Computers cannot work with continuous data. So we need to digitise it!


How does conversion occur?

If we are converting an analogue sound to a digital sound, we use an Analogue to Digital Converter (ADC). The image below illustrates converting analogue sound received on a microphone into a digital signal.

If we are converting a digital sound to an analogue one, we use a Digital to Analogue Converter (DAC). The image below illustrates converting digital audio into a signal that can be played by a speaker.

In most general-purpose computers, the functions of both a DAC and an ADC for sound are handled by the sound card.


But how does digitisation actually happen?

An ADC uses sampling techniques to convert an analogue signal. This involves: sampling, quantization, and encoding. You don't need to remember exactly how this is done, but it really helps you understand how a real-world signal is changed into something the computer can understand. The DAC will do the same thing but the other way round!


This diagram shows how the sampling, quantization and coding steps result in a digital signal.

Let's break down each step:

  1. Sampling The process of slicing a continuous wave equally is called sampling. A sample is a slither in time so, the time axis is divided into fixed intervals of T. The amount of samples taken for every second is called the sample rate. This smaller the T the higher the sample rate which tends to result in a higher quality sound. Then, the reading of each value of analogue signal is taken for each time interval, the vertical arrow on diagram above.

  2. Quantization The process of converting a sample into a discrete value is called quantization. Signals are categorised into a fixed number of intervals. To explain, each interval of the same size is assigned a number. In this example we have split the range from 0 0 0 to 1 1 1 so 8 levels; 0 0 0, 0 0 1, 0 1 0, 0 1 1, 1 0 0, 1 0 1, 1 1 0 and 1 1 1. Note that in this case to represent each and every sample in binary we need 3 bits. Therefore every sample taken has a 3-bit resolution. Each sample falls in one of the intervals and is assigned a number.

  3. Coding The process of representing quantized values digitally is called coding. In our example, eight quantizing levels are used. These levels can be coded using 3 bits if the binary system is used, so each sample is represented by 3 bits. If we string all the samples together in their 3-bit format, we get the digitized sound: 1 0 0 1 1 0 1 0 1 0 1 1 0 0 1 0 0 1 0 1 0.


159 views0 comments

Comments


bottom of page