What parallel compression is and how to set it up
Parallel compression mixes a heavily compressed copy of a sound under the dry original. The dry signal keeps the punch of every hit, and the squashed copy fills in weight and detail between them. In Live, send the part to a return track with a hard compressor, or turn down the compressor's Dry/Wet.
| Stage | Detector | Ratio | Attack | Release | Gain reduction | Threshold, peaks at −6 dB | Dry/Wet |
|---|---|---|---|---|---|---|---|
| Smash a copy on a return | Peak | 8:1 to 10:1 | 1 to 5 ms | 50 to 100 ms | 10 to 15 dB | −20 dB | 100 % |
| Blend with Dry/Wet (optional) | Peak | 4:1 to 10:1 | 1 to 10 ms | 50 to 150 ms | 8 to 12 dB | −18 dB | 30 % |
This page shows 2 ways to set up parallel compression: smash a copy on a return and blend with Dry/Wet.
Pick one. Each way does the whole job on its own, and the first one is the starting point.
Across them the ratio runs from 4:1 to 10:1 and the gain reduction targets from 8 to 15 dB. Set every threshold last, by watching the gain reduction meter, not by copying a number.
The stages, in order
Smash a copy on a return, 8:1 to 10:1
Drums or vocals need weight, but more compression kills the punch. A return track with a hard compressor squashes a copy, and the dry track stays untouched.
Start with the Peak detector, a ratio of 8:1 to 10:1, an attack of 1 to 5 ms and a release of 50 to 100 ms. The fast attack catches each peak before it gets past. Set the release so the meter falls back to zero before the next note or hit, or the next one starts already turned down and loses its punch.
Lower the threshold until the meter shows 10 to 15 dB of gain reduction on the loudest parts. If the part peaks near −6 dB, that is a threshold of about −20 dB.
Blend with Dry/Wet, 4:1 to 10:1
You want parallel compression on one track without a return. Turns down the compressor's Dry/Wet so the dry signal passes under the squashed one.
Start with the Peak detector, a ratio of 4:1 to 10:1, an attack of 1 to 10 ms and a release of 50 to 150 ms. The medium attack keeps some of the front edge while it still holds the peak. Set the release so the meter falls back to zero before the next note or hit, or the next one starts already turned down and loses its punch.
Lower the threshold until the meter shows 8 to 12 dB of gain reduction on the loudest parts. If the part peaks near −6 dB, that is a threshold of about −18 dB.
Turn Dry/Wet down to about 30% so the dry signal carries the punch and the compressed one adds weight underneath.
Use this way in place of the starting point, not on top of it.
The settings in Live 12
CompressorLite, Intro, Standard, Suite
- Model
- Peak
- Threshold
- -18.0 dB
- Ratio
- 8.00 : 1
- Attack
- 1.00 ms
- Release
- 60.0 ms
- Knee
- 0.0 dB
- Makeup
- On
- Dry/Wet
- 30 %
These settings are for the one track method, with Dry/Wet at 30% blending the squashed signal under the dry one. On a return track, set Dry/Wet to 100% and use the return's fader for the blend instead.
Glue CompressorStandard, Suite
- Threshold
- -18.0 dB
- Ratio
- 10
- Attack
- 1
- Release
- .1
- Makeup
- 6.00 dB
- Dry/Wet
- 30 %
Glue Compressor stands in for the Dry/Wet blend stage, because it has its own Dry/Wet knob. At 10:1 with a fast attack it squashes hard, so blend it in slowly.
In short: smash a copy on a return at 8:1 to 10:1 with 10 to 15 dB of gain reduction. Then switch the compressor off and on at the same loudness, and keep it only if the part sounds fuller, not just louder.
Why the blend keeps the punch
A normal compressor has to pick a side. Set it hard and the part gets weight and detail, but the front of every hit gets turned down and the punch goes. Set it gently and the punch stays, but the part doesn't get much heavier. Parallel compression gets you both, because you keep two copies of the sound.
The dry copy stays exactly as it was, with every hit at full strength. The second copy goes through a compressor set very hard, with a fast attack, a high ratio and 8 to 15 dB of gain reduction. On its own that copy sounds flat and squashed. The quiet parts of the sound, like the ring of a drum, the room and the breath between words, come right up to the level of the loud ones.
When you blend the squashed copy in under the dry one, the loud parts barely change, because the dry copy is already much louder there. The quiet parts get filled in, because the squashed copy is loud exactly where the dry one is quiet. The result is a fuller, heavier sound, and the hits keep their full punch.
You need far less of the squashed copy than you might think. Bring it up from silence until you start to hear the part get thicker, then back it off a little. Most of the time it should be something you feel more than hear. Switch it off and the part should sound thinner, not quieter.
Setting it up on a return track
A return track is the classic way, and it is the starting point on this page.
In Live, add a return track and put a Compressor on it. Set Model to Peak, the ratio to 8:1 to 10:1 and the attack to 1 to 5 ms, and lower the threshold until the meter shows 10 to 15 dB of gain reduction. That much gain reduction is what makes it work. Set Dry/Wet on the compressor to 100%, so the return carries only the squashed sound.
Then turn up the send on the track you want to treat. Start with the return's fader all the way down, play the track and bring the fader up slowly. The return's fader is now your blend control.
The return track has a few good points. You can send several tracks to the same return, so the whole kit or a stack of vocals gets squashed together. You can add EQ after the compressor, so the squashed copy only adds what you want, for example cutting its low end so a heavy kick doesn't take over. And you can turn the blend up and down with one fader.
Leave the return's Pre/Post switch on Post, which is Live's default, so the blend stays the same when you change the track's level. Live lines up the timing of the return with the dry track for you, so the two copies stay in phase. Live 12 Lite and Intro have 2 return tracks, and Standard and Suite have 12, so in Lite and Intro you may want to save them for reverb and delay.
Setting it up with Dry/Wet
The second way needs no return at all. Live's Compressor has a Dry/Wet knob, and turning it down blends the dry signal back in under the compressed one, inside the same device.
Put a Compressor on the track and set it up hard, the same way as on a return. Then turn Dry/Wet down to about 30%. The compressed signal now sits under the dry one, as if you had used a return. Glue Compressor in Standard and Suite has its own Dry/Wet knob too, so it can do the same job.
Use this way when you want parallel compression on just one track, or when you have no return to spare. It keeps everything in one place, and you can save it with the track. The trade off is that you can't EQ the squashed copy on its own, and each track needs its own compressor.
Pick one of the two ways for each part. They do the same job, so you don't need both.
What to use it on, and what not to
Drums are the classic use, and still where it does the most. A squashed copy of the kit brings up the room and the ring of each drum, so the kit sounds bigger while the hits stay sharp.
Bass gains from it too. A squashed copy holds the low notes up between plucks, so the line feels heavier without the front of each note going soft.
On a vocal, a little parallel compression brings up the detail and breath and helps the voice sit forward in a busy mix.
Skip it on parts that are already squashed, like a distorted guitar or a dense synth pad. There are no quiet parts left to bring up, so the copy only adds level and noise.
Questions people ask
What should I use parallel compression on?
It works best on parts that need weight but must keep their punch. Drums are the classic use, because a squashed copy brings up the room and the ring of each drum. Bass and vocals gain from it too. Skip it on parts that are already squashed, like a distorted guitar or a dense synth pad, because there are no quiet parts left to bring up.
What is the best compressor for parallel compression?
Any compressor that can squash hard works, because the dry signal keeps the punch. In Live, Compressor on Peak at 8:1 with a 1 ms attack works in every edition, and its Dry/Wet knob lets you blend it on one track. Glue Compressor, in Standard and Suite, has its own Dry/Wet knob too, and at 10:1 it squashes hard.
Why does parallel compression sound so good?
Because you keep both versions of the sound. The dry copy keeps the full punch of every hit, and the squashed copy, with 8 to 15 dB of gain reduction, brings up the quiet ring and detail between them. Blended in under the dry sound, the copy adds weight you feel more than hear, and the hits never get flattened.