Compressors/Limiters – Simplified…. well sort of…

This is my attempt at explaining compressor and limiter technology. I will probably get some bits wrong. Please don’t write in and complain. I’m having a shot at it so just be happy that I’m trying.

Ok

An audio compressor is either a hardware or software device which can automatically reduce the output gain of a signal with respect to its input, by a pre-determined or selectable ratio, once that signal has exceeded a specified threshold value.

Still with me so far?

When using a compressor/limiter device the result is heard as a reduction of dynamic range, allowing the R.M.S. (Root Mean Square) volume being used within a session to be increased without “clipping” or driving over the capacity of the recording (tracking) or playback system.

Got that?

In simple terms a compressor/limiter type device can make audio content sound louder by reducing the difference between the quiet and loud sounds, or dynamic range.

Simples.

Compression is used for a variety of reasons, including noise reduction, emphasis (e.g. to make vocals stand out in a dense mix), ducking, and artistic effects such as “pumping”. These aspects are discussed in more detail below.

It’s probably worth noting that a compressor can be inserted into the signal path at a variety of points, however, it is usually deployed pre-fade and/or on group and mix busses.

Overview of Compressor Technology

Typical compressors work by utilising a variable gain amplifier which is controlled by a signal derived from the amplitude envelope of the signal being processed. Above the threshold level, the control signal acts on the variable gain amplifier to reduce the output gain.

When the compressor is reading signals up to the threshold figure, the variable gain amplifier will operate at unity gain, or a ratio of 1:1 (i.e. for every 1dB increase in input signal amplitude, there will be a 1dB increase in output signal amplitude). Typically the amount of gain reduction, applied above the threshold, will be user selectable and expressed as a ratio. For example, a ratio of 2:1 will produce an output gain of only 0.5 dB for every 1 dB of input gain.

Hang in there.

In some compressor and limiter devices there is a set or fixed threshold with linked input and output gain controls which can be used to push the signal into compression. Most other compressor and limiter devices employ variable threshold and comparator circuitry to achieve the same effect.

Other than the input and output gain controls and the compression ratio there will be several additional parameters. These typically include user variable attack and release times, and sometimes, control for the “knee”. The attack and release times refer to the delay before the gain reduction is applied (attack), and the amount of time the gain reduction remains in operation after the input signal has dropped below the threshold value (release).

Attack and release times can usually be varied from lass than 1ms, to several seconds. The knee parameter determines the rate at which the gain reduction is applied, with a hard knee referring to an almost instant application of the full amount of gain reduction, and a soft knee referring to a more gradual application of gain reduction. Example of both of these types of knee can be seen below in some stolen graphs below.

Many compression/limiting devices have a pre-set knee determined by their design parameters, whilst others enable operator selection of the knee characteristics.

Figure 1. Example: Hard and Soft Knee. (Source: http://www.massenburg.com/cgi-bin/ml/8900ref.html)

Finally, most compression/limiting devices or software includes a gain makeup control to allow adjustment of the compressed output signal to the required level.

Phew, that wasn’t as fun as I’d hoped.

So let’s go again, this time with limiters….

Limiting can be considered as an extreme form of compression, and whilst there is no formal definition, many sound engineers consider compression ratios in excess of 10:1 to be limiting, though ratios can extend up to and including ∞:1 (thanks Kadis, J.).

The main purpose of limiting is to ensure that no programme content exceeds a pre-determined level, and whilst compression is often employed as a sound sculpting tool, limiting is typically (though not exclusively) employed to ensure that the operational parameters of the system (recording/ transmission/ playback) are not exceeded due to high amplitude transients. Thus, it is not unusual to use compression then limiting in series.

So next time you stare at those pretty green and red flickering lights on your Beringher rack you’ll know what’s what.

What?