What Does a Sound Compressor Do?
Most likely, if you have been involved in sound engineering or recording music, you will have noticed that people often sound awful when singing into a microphone. This doesn’t always mean that they are bad singers; some of the best recording artists may find that this happens to them. The problem is that their vocal track sounds bumpy and varies in volume and although this can be reduced with correct microphone use (e.g. positioning, gain level, a popper stopper) it is still evident on the finished recording.
This is where a compressor comes in. This device, now available in its digital form, as well as the older analog version, effectively squeezes the signal into a flatter track. That is to say, it reduces the loudest parts of the recording and raises the volume of lower parts, creating a track with a much more uniform volume level.
However, compressors are not appropriate for every type of music or instrument. For example, compressing a classical music track removes the dynamics and the interest of the piece. Furthermore, compressors are hard to use correctly and their overuse can bring about disastrous results.
The two main variables which compressors use are:
This is the point where the compressor begins to take hold. As such, signals with a volume below this level are left unaltered and those with higher volumes are brought down.
This is the quantity by which the signal is reduced, or compressed, and is given as the ratio of the original volume level versus the threshold level. To illustrate this: A ratio of 1:1 has no effect on the signal, but increasing to 2:1 and over, the difference in the original volume level and the threshold are larger so the compressor works harder to reduce the output signal. If you were to set the compressor to a ratio of infinity to one, it would effectively operate as a limiter – a topic for another day!
Other common controls:
As the input signal arrives at the threshold level you have set, you can configure the compressor to wait a specific time before acting on the signal. As an example, when compressing drum tracks, this would be set at a very low time so that the short peak of a drum hit is caught by the compressor and does not distort.
After the signal has ceased, or crosses the threshold in the other direction, you can again set the compressor to wait before it stops acting on the signal. This varies very much on the instrument and recording environment. If in doubt, it is suggested to use a general release time of around 0.10 seconds.