MM Productions Ltd - av equipment, radio microphones, av equipment hire, PA hire, av sales

Gain Structure - 2 - Noise floor, headroom and clipping

As you probably know, the original sound that was fed into the microphone becomes an electrical signal that is passed from component to component in the sound reinforcement chain.
This electrical signal will have a range of values which will be measured in volts or one of the dB references (see the Q&A on dB). Loud sounds produce signals with more volts than soft sounds.
In a perfect system, each time the signal is passed from component to component, there will be no loss of the signals integrity, and each component will modify the signal in its own way (mixers will mix signals, amplifiers will amplify, equalisers will change the frequency component) without doing anything else.
Of course, this is no such perfect system, and each component will inevitably add a certain amount of two things to the signal - noise and distortion.

Noise
All electrical circuits produce noise. This is the random movement of electrons in the circuit which gives rise to tiny currents and voltages, which in a sound system appear at the speaker as low level hiss. You will hear it if you turn everything up full with no signal. The amount of hiss is called the noise floor.

Imagine a room full of people talking at a low level. Now imagine that you are trying to make out what just one of these people are saying. If they talk at the same level as everyone else, and you are equidistant from all the people, you will not be able to hear what he is saying. He will either have to speak louder than everyone else, or you will have to move very much closer to him so his voice is effectively louder than the rest.
The background voices in this scenario are equivalent to the noise in an electrical circuit, and the wanted voice is the signal. We have to make sure that the lowest of the voltages in our signal are greater than the voltages of the noise floor. The higher they are above the noise floor, the better.

Headroom and clipping distortion
However, we cannot make the lowest voltages too high, for fear that the highest voltages become too much for a component and the signal is 'clipped', that is the peaks are flattened because the units power supply cannot provide enough voltage.

As you can see here, this pure tone signal would sound distorted, and harmonic effects from this clipping may even, at high levels, damage loudspeakers.

Now, because the meters on equipment such as mixing desks and the led lights on amplifiers tend to give a reading of the average signal rather than the instantaneous peaks, if you are setting gain by using these meters, you must leave enough room for signal peaks. The room that you leave is called the headroom, and will vary depending on the type of program material going through the system.
NB it is vital that you do not allow clipping in digital systems (such as the O3D mixer or the KT 716 delay) as digital clipping always sounds terrible. Analogue equipment such as the Soundcraft mixers we supply can tolerate a certain amount of clipping distortion.

< previous page - - next page >