You compress your drum bus with a 4:1 ratio and a fast attack. The drums get tighter. They also lose the snap that made the chorus feel huge. You back the threshold off. Now they sound thin again.
Twenty minutes in and you are stuck choosing between punch and control.
There is a way to keep both. It is called parallel compression.
Parallel compression lets you have the controlled, dense sound of heavy compression and the natural transients of the dry signal at the same time, by blending the two together.
By the end of this post you will know exactly when to reach for it, how to set it up on drums, vocals, bass, and guitar, and the mistakes that quietly add mud to a mix when it is used wrong.
This post is part of the Complete Guide to Audio Compression.
TL;DR
- Parallel compression = blend a heavily compressed copy of a signal with the dry original.
- What it solves: punch and density without losing transients.
- Best on: drums, vocals, bass, sometimes electric guitar.
- Skip on: signals that are already squashed (programmed 808s, synth pads) or recordings where transparency matters more than impact.
Keep reading.
By the end of this post you will have specific attack, release, and ratio targets for each common source, plus a mistakes checklist that catches the issues that quietly turn parallel compression into mud.
What Is Parallel Compression?
Parallel compression (also called New York compression) sends a signal down two paths at once. One path stays untouched.
The other path gets aggressive compression, sometimes 10:1 or higher with a fast attack, that crushes the dynamics into a flat, sustained version of the original.
Both paths arrive at the mix bus and play back together.
You hear the dry signal first. The compressed signal sits underneath, adding body, density, and apparent loudness. Because the dry path is intact, the transients that give the source its impact still come through.
That is the whole trick. Heavy compression for energy. Dry signal for punch. Blend to taste.
Why Parallel Compression Beats a Single Compressor
A single compressor on a track has to make a tradeoff. Compress hard enough to add density and you flatten the transients. Go gentle enough to preserve transients and you barely hear the compression at all.
Parallel compression skips the tradeoff.
The compressed copy can sit at 10:1 with 12 dB of gain reduction without anyone caring, because the dry signal is doing the punch work. You get four practical wins:
- Preserved transients. The drum hits, vocal consonants, and pick attacks all stay sharp.
- Added body. The compressed bus fills in the gaps between transients, making the source feel bigger without making the peaks louder.
- Apparent loudness without crest factor loss. The track sounds louder and denser, but the peak-to-RMS gap (your crest factor) on the dry signal stays intact.
- Mix glue without flattening. A parallel bus across a drum kit can glue the elements together while leaving each individual hit punchy.
The shorter version: parallel compression is what you reach for when you want compression to add something instead of remove something.
Want to see how much parallel compression your track actually needs? Drop your WAV file into the Compression Analyzer.
Parallel Compression Settings: Where to Start
These are the starting-point settings for the compressor on the parallel bus. The dry channel stays untouched. Adjust from here based on the source and the song.
Ratio
Aggressive. 8:1 to 20:1 is the working range for parallel work. The compressor on the parallel bus is supposed to crush the signal. Subtle ratios like 2:1 or 3:1 do not give you the density that makes the technique worth doing in the first place. Save those for the dry channel if you need them.
Attack
Fast. 0.1 to 5 ms on drums and percussion. The whole point is to clamp down on transients in the parallel bus so the compressed copy sounds dense and sustained. The dry channel preserves the punch. For vocals and softer sources, you can move up to 5 to 15 ms if a fully crushed compressed copy sounds artificial in the blend.
Release
50 to 150 ms for rhythmic material. The release should let the compressor recover before the next transient hits. Too slow and the compressed bus pumps in a way that sticks out. Too fast and you get distortion on sustained content. For bass and longer sustained sounds, push to 150 to 300 ms.
Gain Reduction
Aim for 8 to 15 dB of gain reduction on the parallel bus during the loudest parts of the source. This is heavier than you would ever apply to a dry channel. That is the point.
Knee
Soft knee for vocals and acoustic sources. Hard knee for drums where you want the compressor to grab the transient cleanly the moment it crosses the threshold. If you are not sure, start soft and switch to hard if the compressed bus sounds smeared.
Blend
Bring the parallel bus up until you can clearly hear it adding to the dry signal, then back off 2 to 3 dB. If you can solo the compressed bus and it sounds dense and sustained, the settings are right. If it sounds thin or chattery, your attack is too slow or your ratio is too low.
Settings by Source: Drums, Vocals, Bass, Electric Guitar
The starting points above work, but each source has its own quirks. Here is what to adjust on the four sources where parallel compression earns its keep.
Drums
Send the full drum bus to a parallel channel. Compressor at 10:1 to 20:1, attack 0.1 to 1 ms, release 80 to 120 ms, aiming for 12 to 18 dB of gain reduction. Blend until the kit feels bigger and more sustained without the dry transients losing their crack. The classic application of the technique.
Vocals
Use a slightly less aggressive ratio: 8:1 to 15:1, attack 5 to 10 ms, release 40 to 100 ms, gain reduction 10 to 15 dB. Blend the parallel bus low, typically 8 to 15 dB below the dry vocal. The goal is to add density and sustain to the lyrics without the compressed copy becoming audibly squashed.
Bass
Bass is already low-crest by nature, so go a little gentler: ratio 6:1 to 10:1, attack 5 to 20 ms, release 150 to 300 ms, gain reduction 6 to 10 dB. The longer release matches the sustain of bass notes. The slower attack lets the pick or pluck attack come through on the dry channel. Blend until the bass feels glued and consistent without the low end becoming boomy.
Electric Guitar
Mostly useful on rhythm guitar where you want the part to feel thicker without losing the strum dynamics. Ratio 8:1 to 12:1, attack 10 to 20 ms, release 80 to 150 ms, gain reduction 8 to 12 dB. Lead guitar usually does not need parallel compression, since the natural pick dynamics and phrasing carry the part.
How to Set Up Parallel Compression in Your DAW
Three setups work. Pick the one that fits your DAW and your workflow.
- Aux send method (most flexible). Create a new auxiliary or return track. Insert a compressor on it with the settings above. Send the source track (drum bus, vocal, bass) to the aux at unity gain. The aux runs in parallel to the dry source. Adjust the aux fader to control the blend.
- Duplicate track method. Duplicate the source track, set the duplicate to a heavy compression preset, and adjust the duplicate’s fader to taste. Simpler than the aux send, but uses more CPU and breaks if the duplicates ever drift out of sync.
- Mix knob method. Many modern compressors (FabFilter Pro-C 2, Waves CLA-76, iZotope Neutron, plenty of stock plugins) have a wet/dry mix knob. Insert the compressor directly on the source, set it aggressively, then back the mix knob off until the dry signal is dominant. Fastest setup, but no independent EQ on the parallel path.
The aux send method is the standard. Use it whenever you want the option to EQ the parallel bus independently of the dry channel.
Try it on anything: a drum bus, a parallel vocal chain, a bass DI, a full mix. Launch the Compression Analyzer → About 5 seconds per file, runs entirely in your browser.
Workflow Tips That Save You Time
- Set the parallel bus first, then the blend. Solo the compressed bus and dial it in until it sounds dense and sustained on its own. Then unmute the dry signal and ride the blend fader. If you try to balance both at once you will spend an hour chasing your tail.
- Save instrument-specific presets. Drum parallel and vocal parallel use very different attack and ratio numbers. Build a starting-point preset for each so you are not rebuilding from scratch on every session.
- Automate the blend, not the threshold. If you want more parallel impact in a chorus, automate the parallel bus fader. Automating the threshold inside the compressor changes the character of the sound, not just the amount.
- Always level-match before A/B. The parallel bus adds apparent loudness. Match levels with the bypass before deciding if it actually sounds better. Otherwise you are comparing “louder” to “quieter,” not “with parallel” to “without.”
- Listen on small speakers. Heavy parallel compression can sound great on monitors and turn into mud on a phone. Reference both early.
Common Parallel Compression Mistakes
1. Overcompressing the Parallel Bus
The parallel bus is supposed to be heavily compressed. That is the whole design. But “heavily compressed” has a ceiling. Past 18 dB of gain reduction the compressed copy starts pumping audibly, distorting on sustained content, and adding harshness instead of body. If you need more density, raise the parallel bus fader. Do not push the compressor harder.
2. Clashing Frequencies
Heavy compression exaggerates everything in the signal, including frequencies you did not want emphasized. A drum parallel bus often gets boomy because of low-frequency bleed from kick into the overheads. The fix is to high-pass the parallel bus at 200 to 300 Hz so the compressed copy adds energy in the mids and highs while the dry signal handles the low end.
3. Blindly Following Settings
The numbers in this post are starting points. A live drum kit recorded in a small room and a programmed trap kit need different ratios, attacks, and blends. Listen to the source and adjust. The settings are a compass, not a destination.
4. Skipping A/B Testing
Mute the parallel bus and listen. Unmute it and listen. Repeat. If you cannot clearly hear what the parallel bus is contributing, either it is doing nothing or it is masked by something else in the mix. Find out which before you commit.
5. Skipping Critical Listening on Other Systems
Parallel compression that sounds tight on monitors can turn into mud on phone speakers and laptop speakers. Reference the mix on at least three playback systems before you call it finished. The difference is often a few dB on the parallel bus blend.
Ready to stop guessing your settings? The Compression Analyzer reads your track and recommends settings that fit your actual source, not a generic preset.
Parallel vs Traditional Compression
Traditional (serial) compression sits directly on a track and processes the entire signal. Whatever the compressor does to the dynamics, it does to everything you hear.
That is great for taming peaks, leveling out a vocal, or controlling a bass that is jumping all over the place.
Parallel compression keeps the original dynamics intact and adds a heavily compressed copy underneath. You get density without giving up transients.
The tradeoff is more setup time, more CPU, and a blend control that has to be managed.
Use traditional compression to control a signal. Use parallel compression to enhance a signal. Most full mixes use both: traditional on individual channels, parallel on busses where you want size and impact.
Frequently Asked Questions
Is parallel compression worth it?
Yes, on the right sources. Drums, vocals, and bass almost always benefit from a parallel bus when you want size and density without flattening the transients. On already-compressed sources like programmed 808s or commercial sample loops, parallel compression adds little because there are no transients left to preserve. Use it when you want the compression to add something, not when you need it to fix something.
Does parallel compression go before or after EQ?
EQ first, then parallel compression. Shape the tonal balance on the dry signal, then send the EQ’d signal to the parallel bus. If you EQ after the compressed bus is already blended in, any tonal change ends up affecting both paths in ways that are hard to predict. The exception is high-passing the parallel bus itself at 200 to 300 Hz to keep low-frequency mud out of the compressed copy.
Is parallel compression the same as bus compression?
No. Bus compression applies a single compressor to a group of tracks routed to the same bus, and everything on that bus gets compressed together. Parallel compression splits a signal into two paths, compresses one heavily, and blends them back together. You can apply parallel compression to a bus (a parallel drum bus is the classic example), but the technique itself is about the dual-path blend, not the routing.
Should parallel compression be in mono or stereo?
Match the source. A mono lead vocal, kick drum, or bass works fine with mono parallel compression and avoids any phase weirdness. A stereo drum bus or a stereo synth pad needs stereo parallel compression to preserve the image. Mono compression on a stereo source collapses the stereo field and is almost never what you want.
What ratio should I use for parallel compression?
Aggressive. The working range is 8:1 to 20:1. Lower ratios than 8:1 do not produce the density that makes the technique worth setting up in the first place. The dry channel is doing the punch work, so the parallel bus is free to crush the dynamics. If you want a gentler effect, lower the parallel bus fader rather than the compressor’s ratio.
Can I use parallel compression on the master bus?
Yes, but with caution. Parallel compression on the master bus can add glue and apparent loudness to a finished mix, but it also affects every element at once, which can mask the very transients you usually want to protect. If you go this route, use gentler settings (ratio around 4:1 to 6:1, gain reduction 3 to 6 dB) and blend the parallel bus low. For most projects, parallel compression is more useful on individual sources or busses than on the master.
The Bottom Line
Parallel compression is the technique you reach for when a single compressor cannot give you both punch and density at the same time. Set the parallel bus aggressively. Blend it underneath the dry signal. Match the settings to the source instead of copying a preset.
And if you want to know what those settings should actually be on your specific track, drop the file into the Compression Analyzer. It reads peaks, RMS, and crest factor, then hands you the attack, release, and ratio that fit your source. About 10 seconds per file, free for 3 analyses a day.
If you want to go deeper on specific applications, these are worth your time:
- Parallel Drum Compression: Step-by-Step
- Parallel Compression on a Vocal
- Mix Bus Compression Settings
Practice and experimentation are key. Trust your ears, use the numbers as a compass, and you will nail compression faster.