What is parallel compression?
Parallel compression is a mixing technique where you blend a heavily compressed copy of a sound with the uncompressed original. The dry signal keeps its natural dynamics and transients. The compressed copy adds weight, density and sustain underneath it.
It’s also called “New York compression”. Its net effect is similar to upward compression: quiet detail rises while the transients stay intact.
Why use it
- It keeps the original dynamics. You get control and punch without squashing the performance.
- It adds weight without obvious compression. The dry track still sounds like itself.
- It suits sources that need to stay natural. Heavy direct compression on a vocal can rob it of nuance, so a blend is often gentler.
Parallel compression is most common on drums and vocals. It also works on bass and, used subtly, on a full mix bus.
How to set it up
There are two common methods. Both work in any DAW.
Aux send method
- Create an aux track or bus in your DAW.
- Send the source track (or a group of tracks) to the aux.
- Put a compressor on the aux and compress it hard.
- Bring the aux fader up slowly, underneath the dry track, until it adds weight without taking over.
The aux is more flexible. You can add a second compressor, EQ the copy or add saturation to it.
Wet/dry knob method
Put a compressor directly on the track, drive it hard, and use the compressor’s mix (wet/dry) knob to blend in the squashed signal. This is quicker, but you have fewer options for shaping the copy.
Delay compensation
Keep your DAW’s plugin delay compensation (PDC) switched on. If the dry and parallel paths aren’t aligned, you’ll hear comb filtering, which sounds thin and phasey. Avoid putting high-latency or linear-phase processing on only one of the two paths.
Starting settings
These are starting points, not rules:
- Drum bus: a ratio of about 4:1 to 10:1 (higher if you want more), a fast attack, a medium-fast release and 10 to 20 dB of gain reduction.
- Vocals: gentler, around 4:1 with 6 to 10 dB of gain reduction on the parallel copy.
- Blend: often about 10 to 30 percent of the compressed signal, set by ear.
Some engineers reach for a classic 1176-style compressor in its “all buttons” mode for a popular parallel drum sound.
Parallel compression on drums and vocals
Drums
Send the kick, snare and toms more than the overheads. Heavy parallel compression on cymbals can make the compressor pump as they ring, so keep the overhead send small. Once the copy is set up, EQ it to add low end or top end.
Vocals
A parallel copy brings up quiet details and breaths, so the vocal stays present and consistent while the dry track keeps its natural feel. Keep the blend low, so the vocal sounds fuller rather than compressed.
Common mistakes
- Louder doesn’t mean better. The compressed copy adds level, which can fool your ears. Level-match the blend, then switch the parallel track on and off. When it’s off, the mix should sound weaker, not quieter.
- Judging the copy on its own. A heavily squashed copy often sounds bad solo. That’s expected. Judge only the blend, in context.
- Too much blend. It brings back the flattened sound you were trying to avoid. Keep the parallel signal below the dry one.
- Forgetting automation. If you automate the dry track’s volume, the parallel path won’t follow unless you automate it too. Keeping the two in step preserves the balance.
Where it fits with other compression
Parallel compression doesn’t replace ordinary compression. Use normal compression to control peaks, and use the parallel copy to add weight. A de-esser placed after compression helps if the squashed copy makes sibilance more obvious. For a single vocal track, see vocal compression settings that work and what a de-esser is and how to use one.
For a compressor to use on the parallel bus, see the best free compressor VST plugins for home mixing. The send-based routing is similar to a reverb setup, explained in reverb sends vs inserts.