1.2 Memory and Storage

On this page, we will look at how a sound can be recorded and converted to a form that a computer system can store.

The continuously changing sound signal has to be sampled many thousands of times a second. The microphone produces a signal and an electronic circuit converts the signal to a number at very precise and regular intervals. The amount of signals that are recorded a second is called the sampling frequency and may be as high as 300,000 times a second but can also be as low as 5000 times a second.

To replay the sound, these digital values are used to control the level of an electrical signal that moves a loudspeaker cone at precisely the correct frequency (the sampling frequency). The loudspeaker cone move back and forwards which vibrates the air and causes a sound pressure wave to travel through the air in exactly the same way that the original sound did.

When the signal is converted to a number to be stored in a computer, the number values need to be converted to a binary value. If we use one byte to store each sampled value, then it is possible to store 256 different values. This is known as the bit depth of the sample. An eight bit bit depth will create a reasonable sound quality but when the sound is replayed, it will be obvious that the sound quality is not as good as the original. If we use two bytes instead, the a 16 bit value can store 65536 different values which will produce a much better sound that the human ear will not be able to tell from the original sound.