Audio: normalisation (0:04)

Figure: normalisation

normalisation plays a ringtone twice. The normalisation figure shows the composite waveform. The first ringtone is normalised to -1dBFS and is shown in the left-hand half of the waveform, the second ringtone is normalised to -9dBFS and is shown in the right-hand half of the waveform. There is a noticeable difference in volume between the two sounds.

Normalisation is a process of increasing or reducing amplitude to a target level.

Normalisation is a safety feature. It prevents a sound from being too loud and distorting during playback. Setting a target amplitude level, a norm, during recording, ensures that the output will not exceed this level.

In digital audio recording, amplitude that exceeds the norm may cause clipping. Clipping is unwanted distortion and is to be avoided.

Headroom is the difference between peak amplitude and normal amplitude. What is considered a reasonable amount of headroom is a matter of debate. The European Broadcasting Union, for example, specifies a maximum level of -9dBFS, which allows for 9dB of headroom.

All the music in the guide is recorded with between 3dB and 10dB of headroom. The final version is normalised to -1dBFS.

It is advisable to DC offset digital audio at the same time as it is normalised. Digital samples, especially those downloaded from the internet, occasionally include unwanted noise which distorts the waveform and DC offset is a process for removing this bad noise.