There are few audio processing techniques that mange to confuse the home studio owner more than compression. So let us try and take the mystery out of it and shed some light on its basic applications.
Simply put, compression is the process of decreasing the difference in level between the loudest and quietest parts (the dynamic range) of an audio signal. So why might we need to do this? Lets consider the example of a vocal recording where the vocalist sings some of the words noticeably quieter than others. If you where to set the vocal level where the majority of the words sit nicely in your mix, the quieter words are going to disappear behind the music. Conversely if you fade the vocal up to where the quieter words can be heard, then the rest of the vocal is going to blow your ears off.
Compressors were invented to provide a solution to this problem. By reducing the dynamic range between the loudest and quietest words in our vocal example it becomes easier to set a vocal level that works for the vocal as a whole.
Lets take a look at some common compressor controls.
- Threshold – sets what level the signal must be before the compressor starts working (kicks in).
- Ratio – sets how much compression is applied. E.g. if the compression ratio is set for 2:1, the input signal will have to cross the threshold by 2 dB for the output level to increase by 1dB.
- Attack – how quickly the compressor kicks in after the signal exceeds the threshold.
- Release – how fast after the signal dips below the threshold the compressor resets.
- Knee – sets how the compressor reacts to signals once the threshold is passed. Hard Knee settings mean it acts on the signal the moment it exceeds the the compression threshold, and Soft Knee means the compression kicks in more gently as the signal reaches the threshold. Soft knee settings are typically more musical although hard knee settings can be desirable for certain applications.
- Make-Up Gain –lets you Boosts the resultant audio after compression, as compression can reduce the signal significantly.
- Output – allows you to boost or cut the level of the signal output from the compressor.
How To Set Up a Compressor
1. Whether you are using a hardware or software compressor the settings are going to remain the same. So grab your compressor of choice and insert on the channel you want to compress.
2. Set the threshold until the signal is pushing over the threshold and triggering the compressor. Look for something like 6 dbs of gain reduction so we can really hear whats going on .
3. Set the Ratio according to the material. Bass guitars sound good at 4:1, drums at 2:1, vocals also at 2:1 and electric guitars anywhere from 2:1 to 6:1.
4. The Ratio and Threshold work in unison. Adjust them together and see how they affect the output.
5. The attack and release controls decide how the compressor will react. E.g. slow attack, slow release would be useful for a snare drum by allowing some of the drums transient past before the compressor starts to act on the signal.
6. If available, choose between Hard and Soft Knee. Hard Knee can work well on drums where as Soft Knee can be better suited for vocals or more melodic instruments.
7. Adjust the Make up gain to compensate for the decrease in signal level.
All suggested settings should be taken with a pinch of salt, no two audio signals are the same so some experimentation is necessary.
Compression Audio Sample
Lets take a snare sample a loop and apply different attack times so we can hear what it sounds like in practice.
Snare Uncompressed: In the following samples I have set the threshold to -13db, ratio at 6:1 and release at 200ms.
Snare 1ms: With an attack time of 1ms the compressor is reacting very fast and blunting or softening the initial attack transient.
Snare 10ms: With 10ms attack the compressor is allowing the initial attack through uncompressed which accentuates this portion of the sound (it sounds more clicky!).
Snare 50ms: A greater portion of the attack is now passing through the compressor before it reacts which results in gain reduction that mostly affects the decay.
Snare 100ms: At 100ms the majority of the signal has passed through the compressor before it reacts, resulting in just a subtle reduction in level on the signals decay.