TUTORIAL: Demonstrations of compression on female vocal
The singer is Audio Masterclass Featured Artist, Chilli Gold.
As you will hear, the vocal is already clear and well recorded. Notice the small mouth noises, ventilation noise in the background, ambience, a singer rehearsing in another room at one point, and the assistant engineer's comment at the end. All of these will reveal interesting characteristics in the compression employed.
Let's start with the default setting of the DigiRack plug-in. Clearly Avid will have thought long and hard about which settings will be generally most useful...
As you will hear, the compressed version is lower in level than the original. This is normal. Make-up gain is required to bring the peaks back to where they were in level approximately. This will have the effect of bringing the quiet sounds up in level. As with all compressors, this increases the noise level, whatever the source of noise in the original signal.
The only control that is changed is the red Gain control on the right...
Here we have changed the compression ratio to a lower value. To achieve the same amount of compression (so that the amount of gain reduction is approximately the same), the threshold has been lowered.
The relevant controls are the gray Ratio control, and the yellow Threshold control...
In the next example, the compression ratio is increased. The threshold has been raised to keep the amount of compression approximately the same...
Here, the settings have been put back to their default positions, except for the Release control, the value of which is lowered. Setting a short release time increases the perceived strength of the compression effect. However, the compression may be more noticeable. If it is required that the compression is unnoticeable, a long release time is preferable.
The relevant control here is the gray Release control at bottom center...
In this example, a long release time has been set...
In the above example, the compression is certainly very smooth. In fact it is almost unnoticeable. The reason for this is that there is hardly any active compression going on. Compression only occurs when the gain reduction (GR) meter is changing. Setting a very long release time reduces the amount of active compression almost to zero. If the gain reduction meter shows a fixed or very slowly changing reading, then what you are hearing is almost the same as if you had lowered the fader by that number of decibels. Try it.
The knee control below sets the rate at which the compression effect starts to 'bite' as the signal rises above the threshold and continues to rise. In general knee ranges from 'hard' to 'soft'. A hard knee means that the transition from not-compressed to compressed is sudden. A soft knee means it is gradual. In the compressor shown here, knee is measured in decibels, which we can presume to mean the range of decibels over which the transition from not-compressed to compressed takes place.
You would set a softer knee if you wanted the onset of compression to be unnoticeable. The effect is very subtle and you will have to compare this example carefully with the example made at the default settings (above)...
For most purposes the attack time can be set to 10 milliseconds or so. You would only need to adjust it if the initial start of each sound wasn't sounding right. Here, you can hear that the compressor is slow to respond.
All the controls are at their default settings, apart from the gray Attack control in the center...
It might be thought that a short attack time would be better, so that the compressor can respond immediately. However, this often distorts the initial transient and can sound unpleasant. Here is a very short attack. All of the other controls, including make-up gain, are in the same positions as the previous example so that you can compare them directly...
Finally in this section we will hear the Focusrite Liquid Mix set to emulate the classic Fairchild 670 variable-mu vacuum tube compressor. The result is rather different to anything the DigiRack plug-in can achieve...
Download the audio files for this tutorial...