Parallel compression is the technique of blending a heavily compressed copy of a signal with the original, uncompressed (dry) signal. Instead of compressing a track directly, you mix in a squashed duplicate underneath it. The result keeps the natural punch and transients of the dry sound while the compressed copy fills in body, sustain and density. It is one of the most useful tricks for making drums, vocals and whole mixes sound bigger without losing life.
What is parallel compression doing?
Normal (series) compression turns the loud parts down and then brings the whole signal back up, which can flatten transients and suck out energy if pushed hard. Parallel compression takes a different route: it leaves the dry signal alone and adds a separate, heavily compressed version on top. Because the dry track still carries the original peaks, you keep the punch — while the compressed layer raises the quiet detail and adds weight. Blend the two and you get loudness and density plus dynamics, which is hard to achieve any other way. If the distinction between the untouched track and the processed copy is fuzzy, our dry vs wet signal explainer makes it concrete.
It is sometimes called “New York compression,” after the technique’s popularity in New York mixing rooms, especially on drums. If the basics of compression are new to you, our EQ and compression fundamentals guide is the place to start.
Why use parallel compression?
- Punch with power: drums hit hard (dry transients) but also sound fat and full (compressed layer).
- Detail without squashing: quiet parts of a vocal or room get raised so nothing disappears, while the loud parts stay natural.
- Loudness without obvious pumping: you can add a lot of density while still controlling how audible the compression is, just by adjusting the blend.
- Control by feel: the wet/dry balance is intuitive — turn the compressed copy up for more effect, down for less.
How to set up parallel compression
There are two common ways to do it in any DAW:
- Aux/send method: create an aux (effects) bus, put a compressor on it, and send your track to it. Compress the aux hard, then blend that bus underneath the original track. This is the standard approach and easy to automate; if routing trips you up, see how to set up sends and returns in a DAW.
- Duplicate track method: copy the track, compress the copy heavily, and mix it under the original.
For the compressor settings on the parallel copy, go aggressive: a high ratio (around 4:1 to 10:1), a fast attack, a medium release, and enough threshold to see 10–20 dB of gain reduction. Then start with the parallel level all the way down and slowly raise it until the source gets fuller and more solid. The moment it starts to sound unnatural or lifeless, back off. If the controls themselves are unfamiliar, our guide to using a compressor walks through each one.
Tip: if your compressor has a built-in wet/dry “mix” knob (like FabFilter Pro-C 2), you can do parallel compression on a single track just by turning the mix below 100 percent — no separate bus required.
Dialling in the compressed copy
The whole point of parallel compression is that the compressed layer can be pushed far harder than you would ever dare on the dry track, because the dry signal is still doing the heavy lifting for transients and dynamics. That changes how you approach each control.
- Ratio: high is fine here. A ratio of 4:1 is a sensible starting point, and 8:1 or 10:1 (or even limiting) is perfectly usable because the squashing is only ever heard in the blend, not on its own.
- Attack: a faster attack clamps down on the initial peaks of the compressed copy, leaving the dry transient to define the hit. This is what gives you that “punch on top of a wall of body” character on drums.
- Release: aim for a release that lets the compressor recover roughly in time with the material — too fast and it can sound grainy, too slow and the layer stays ducked and loses energy. On rhythmic material, set it by ear so the level pumps gently in time, not against the groove.
- Gain reduction: 10–20 dB is normal on the parallel channel. Watch the meter, but trust the blend more than the number.
A useful trick when first learning is to solo the compressed bus and push it until it sounds genuinely crushed and a little ugly, then bring the dry signal back and lift the parallel level from silence. The “ugly” copy stops sounding ugly the moment it sits beneath the natural track — it just adds glue and weight.
Where parallel compression works best
- Drums: the classic use. Parallel compression on the drum bus or just the room/overheads makes a kit sound huge while keeping the transient snap, and it slots neatly into the workflow in our how to mix drums guide.
- Vocals: adds density and helps a vocal sit consistently in the mix without over-compressing the lead directly. Pairs well with the chain in our how to mix vocals guide.
- Bass: evens out level and adds sustain while keeping the attack.
- Full mix / mix bus: used gently, it can add glue and loudness to an entire mix.
Common mistakes to avoid
Parallel compression is forgiving, but a few habits will rob it of the very thing it is meant to deliver.
- Adding too much: if you keep raising the parallel level, you eventually drown the dry signal and end up with the same flat, squashed sound you were trying to avoid. The effect should be felt more than heard — pull it back until you notice it leaving, then nudge it up a touch.
- Forgetting to EQ the parallel bus: a heavily compressed copy can bring up boxy low-mids or harsh top end. A gentle high-pass to remove rumble, or a small cut where the layer sounds muddy, keeps the blend clean without changing the dry tone.
- Ignoring gain staging: if the compressed copy is much louder or quieter than the dry signal before you blend, your fader moves become unpredictable. Match levels roughly first so the blend control behaves smoothly.
- Using it to fix a bad balance: parallel compression adds body and density, but it will not rescue a poorly recorded or badly balanced track. Get the source sounding right first, then reach for it as a finishing move.
Watch the phase
Because you are mixing two versions of the same signal, phase matters. Some compressors introduce a tiny delay (latency), which can cause thin or hollow sound when the wet and dry are combined. Modern DAWs usually apply plug-in delay compensation automatically, so the aux method stays in phase. If a parallel blend sounds thin instead of fuller, check that delay compensation is on, or use the single-track mix-knob method to avoid the issue entirely.
Parallel compression is a core mixing move worth practising. For the wider context, explore the mixing and mastering hub and our beginner’s guide to mixing your first song to see where it fits in a full mix.
Frequently asked questions
What is the difference between parallel and serial compression?
Serial (normal) compression processes the whole signal and replaces the original. Parallel compression keeps the original dry signal untouched and blends a separately, heavily compressed copy underneath it, so you keep natural transients while adding density.
How much parallel compression should I use?
Start with the compressed copy turned all the way down, then raise it until the sound gets fuller and more solid. Stop before it sounds squashed or lifeless. The right amount is the most you can add while the source still feels natural.
Can I do parallel compression in any DAW?
Yes. Every DAW lets you create an aux/send bus with a compressor on it, or duplicate a track and compress the copy. Many compressor plug-ins also include a wet/dry mix knob, which lets you do parallel compression on a single channel with no routing at all.
Should I EQ before or after the parallel compressor?
Either works, but they do different jobs. EQ before the compressor changes what the compressor reacts to — for example, high-passing the input stops low-end rumble from triggering gain reduction. EQ after the compressor shapes the finished tone of the blend. A common approach is a gentle high-pass before and a small tone correction after.
