Before I found compression, my life was empty and void. Now that I have it, I feel so alive and…ok, so compression isn’t THAT amazing but if you’ve got it, you should learn how and when to use it.
What is compression?
Audio compression is the process of reducing the dynamic range of a sound. This compression occurs when the volume level signal exceeds a specified level. In practical terms, when a singer decides to belt out the chorus, instead of jumping for the fader, the compressor does the work for you.
When should compression be used?
I prefer to use compression on any channel in which the volume could potentially spike quickly and noticeable unbalance the mix. For example, if an electric guitarist changes pedal effects during a song then there’s the potential for the new pedal effect to have a higher gain and thus become massively louder that everything else in the mix. It’s ok (and good) to have volume fluctuations in a mix; a louder vocal can give the feeling of more of an emotion, for example.
These are instances where volumes fluctuations can occur but will be minimal and likely acceptable in the mix. Or, when a volume increase does occur, you can handle it with a fader change.
Let’s look at where it helps:
- Snare drum
- Kick drum
- Pedal-controlled guitars
- Bass guitar
In the case of the guitars, the musician can easily change to the wrong patch cable or hit their volume pedal and blast the congregation to the back of the sanctuary. For snare and kick, these are two instruments I find the drummer can…”show their enthusiasm.” And vocals, naturally. But this isn’t to say all vocals should be compressed. Some singers need it and some do not – some don’t vary their volume much and others know to move the microphone away to offset it.
How do I set channel compression?
Before covering that, let’s look at how compression is implemented. A compressor can come on quickly or slowly. It can compressor a lot or a little. It can keep compressing, after the volume drops below the threshold, for a short or long period of time. Each of these properties is configurable for a reason.
Imagine the drummer hits the snare drum twice as hard as normal. The compressor should engage quickly and compress a lot of the audio signal. This amount of compression is viewed as the compression ratio.
Ratios show how loud the signal must be in decibels in order to allow 1 decibel to pass through the compressor. For example, a 1:1 ratio means that for every one decibel that goes into the compressor, one comes out. 2:1 means for every 2 decibels over the threshold (the point you want the compressor to engage), only 1 decibel is passed through. So 2 decibels are compressed into 1.
Back to the snare drum. In this case, a high ratio such as 3:1 or 4:1 might be useful. If the volume is 8 dB higher at the source, with a 4:1 ratio, the audience only hears a 2 dB increase.
The trick to using compression is deciding how much variation in the volume level you want that instrument or singer to have over the mix until it’s time to control it. For example, hearing a vocalist increase their volume during a particular passage of a song can be an excellent means of song arrangement or simply magnifying the emotion of a song at a critical point. The volume increase is good but you don’t want them to get too loud.
I find it helpful to set the threshold last. The different compressor properties all work together for a specific need such as mentioned with the snare drum. By setting these to the need, then when the threshold is altered, what is heard is pretty-close to what you’d want to hear – pending final tweaks.
I set ratios high (4:1, 3:1) for guitar and snare. Then lower for the others, in the 2:1 and 2.5 area. Once this is set, I’ll set the ATTACK rate – how quickly the compressor kicks in. Fast for snare and electric guitar, slower for others. Then comes the release rate. With percussive bass sounds, such as kick drum, they can appear to lose bass content if attacked under 50ms. As much as you’ll be listening to how the volume is affected, be aware of how the tone is affected.
The RELEASE controls how soon after the signal dips back below the THRESHOLD that the compressor stops working. A snare hit would be quick so I’d want a quick release. A vocalist would be more likely to prolong a loud vocal line, so I’d increase the release time so instead of their volume going up, getting compressed, lowering, then they sing the next word and it gets compressed…you get the idea. Think of a long RELEASE time as a predictive time. The more likely the vocalist is going to keep singing loud, the longer the release time.
So far, it sounds like a compressor immediately turns on when the threshold is exceeded. That is…an option. The KNEE sets how the compressor reacts when the threshold is reached or almost reached. A hard knee setting mean it turns on immediately while a soft knee means the compression is more subtle in how it’s first applied. A soft knee means the decibel reduction begins as a curve of varying ratios of compression until the set ratio is reached.
I tend to associate fast attack times with hard knees and slow times to soft knees. But that’s just me.
When it comes to setting the RELEASE, KNEE, and RATIO, it’s important the compressor is working to best minimize the volume increases without the listener knowing it’s working. If they hear volume go up and down and up and down then they are hearing the compressor.
There can also be a make-up gain control which allows you to boost the compressed signal. This is where things get wonky. When a dynamic range is compressed, its overall signal is reduced. If it’s reduced too much then the overall sound doesn’t sit right in the mix.
There’s another option sometimes available called a look ahead control. It’s designed to overcome the problem of being forced to compromise between slow attack rates that produce smooth-sounding gain changes and fast attack rates capable of catching transients. But it comes at a slight, though doubtfully obvious cost. The “look ahead” splits the input signal into two paths and delays one. The real-time audio is what’s evaluated by the compression and the information gained from reading it is applied to the delayed signal which is the one that’s compressed and sent out of the compressor.
Back to threshold
With these different controls set, the point in which the compressor kicks in needs to be established. First thing, know that a compressor should not be working 100 percent of the time. That’s over-compressing the signal and you’ll lose the quality of the signal. It should only work when you need it.
I find it best to set the threshold as the band is practicing. When an instrument or vocal gets too loud, then I’ll set the threshold and listen to the result. If it’s too loud, I decrease the threshold. If it’s not loud enough, I increase it. You should be able to see some sort of indicator of when the compressor is kicking in.
Compression can also be used an effect. I can beef up a snare with a second channel of heavily compressed snare and then I blend the two for the punch and a bit more oomph. That being said, it shows compression can (does) alter the sound of the signal. In instances where it’s used for volume control, it shouldn’t be noticeable. If the vocal characteristics, in the case of a singer, seem noticeably different, then reduce the amount of compression be it at the threshold, ratio, attack, or release.
When do I apply compression?
Usually, effects are added after the initial EQ process. In the case of compression for volume control, it depends. If the compressor is used as a “just in case” effect or in the case where it’s only going to come on a couple of times, I’ll set my EQ’s and then add the compressor. If you’ve got a situation where the compressor is coming on a lot, such as with a vocalist (singing or speaking) that is all over the volume range, then set the compressor first and then the EQ.
For easy review, here are the compressor controls:
- Threshold: When the compression will occur. Measured in dB.
- Ratio: How many decibels required for every 1 dB increase. Examples: Vocals (2:1) and Electric Guitar (6:1).
- Hard Knee/Soft Knee: Hard Knee tells the compressor to kick in once a threshold is reached. Soft knee says, “Start compressing a little when I start to reach the threshold.”
- Attack: Determines how quickly compression will kick in. Generally long attack for speaking/vocals and short for percussive sounds.
- Release: Determines how long compression will occur after the signal drops below threshold. Short on percussion, longer on vocals.
- Output: Gain control for raising the signal after compression. Should be set to compensate for attenuation. Think of it this way, you want to compress but you want to maintain a certain output volume during compression.