Most producers spend years using compressors wrong before something clicks. Either they squash everything until it sounds lifeless, or they avoid compression entirely because they can’t tell what it’s doing. Both make mixes worse.
Once you actually understand what a compressor does and how the controls work together, the whole thing stops being mysterious. This guide walks through compression from the ground up: how it works, the six parameters every plugin shares, the workflow I use in my sessions, and a settings cheatsheet by source.
What Is A Compressor?
A compressor reduces the volume of the loudest parts of a sound. That’s the simple version. When a vocal jumps from quiet to loud, the compressor pulls the loud parts down so the difference between the two is smaller. The result is a more consistent signal that sits better in a mix.
The technical term for the difference between loud and quiet in any audio signal is dynamic range. Compression reduces that range. Loud parts get pulled down. Quiet parts stay where they are, or get pushed up later with makeup gain. The signal gets more controlled, more even, and easier to balance against everything else in the mix.
Three reasons you’ll use a compressor on most tracks:
- Consistency: A vocal that whispers and then belts is hard to mix. Compression evens out the performance so every word sits at a usable level.
- Punch and impact: A compressor with the right attack and release settings can make a drum hit harder by accentuating the transient.
- Glue: Compression on a group bus (drums, vocals, the master) ties separate sounds together so they feel like one cohesive element.
The fastest way to understand what a compressor does is to listen to one in action. Here’s the same vocal take, once raw and once with a moderate compression setting applied.
Notice the jump between the quieter phrases and the louder lines.
Threshold at -22 dB, ratio around 2.5:1, attack 20 ms, auto release. About 3-6 dB of gain reduction on the peaks.
The compressed version sits at a more consistent volume. The quiet phrases feel closer to the louder ones. Nothing about the performance changed. The compressor just made the dynamic range narrower. That’s the entire job.
The 6 Compression Parameters You’ll See on Every Plugin
Every compressor plugin shares the same core controls. Once you understand these six, you can sit down with any compressor and dial in a sound without reading the manual. The visual style and the colour might change, but the controls do the same thing.
The level at which the compressor starts working. Set it at -10 dB and anything louder than -10 dB gets compressed. Anything quieter passes through untouched. Lower threshold means more of the signal gets compressed.
How much compression gets applied to signals above the threshold. A 4:1 ratio means for every 4 dB above the threshold, only 1 dB comes out. The higher the ratio, the more aggressive the compression. Anything above 10:1 starts behaving like a limiter.
How fast the compressor reacts after the signal crosses the threshold. Fast attack catches transients (good for taming peaks). Slow attack lets transients through before the gain reduction kicks in (good for keeping drums punchy).
How fast the compressor stops working after the signal drops below the threshold. Fast release sounds aggressive and pumping. Slow release sounds smooth and natural. Most modern compressors have an “auto release” mode that adapts to the signal.
Controls the transition around the threshold. A soft knee eases the compressor in gradually as the signal approaches the threshold. A hard knee snaps it on the second the threshold is crossed. Soft for musical, hard for surgical.
Compression reduces the overall level of the signal. Makeup gain adds it back so the compressed and dry signals have similar perceived loudness. Most compressors have an “auto gain” option that handles this automatically, but manual control gives more precision.
The level at which the compressor starts working. Set it at -10 dB and anything louder than -10 dB gets compressed. Anything quieter passes through untouched. Lower threshold means more of the signal gets compressed.
If I had to pick the two most important controls on any compressor, they'd be Attack and Release. Threshold and ratio just tell you how much compression is happening. Attack and release determine what the compression sounds like.
A slow attack on a snare keeps the crack but tames the sustain. A fast attack on the same snare kills the crack and makes it sound dull. Learn to set attack and release by ear and everything else falls into place.
The Different Types of Compressors
Not all compressors sound the same. Different circuit designs in hardware and different modelling approaches in plugins give compressors distinct characters. The type you reach for affects the colour and the response, not just how much gain reduction you get.
VCA (Voltage Controlled Amplifier) compressors react quickly and predictably without adding much colour to the signal. They're the workhorses of mix bus and drum bus compression because they keep the energy of a mix intact while controlling peaks.
The SSL G-Master Bus Compressor is the most iconic example, used on more hit records than anyone can count. Plugin versions like the SSL G-Master Buss Compressor or the SSL Native Bus Compressor 2 do the same job inside the box. The API 2500 is another classic VCA-style compressor that adds punch and cohesion without flattening the mix.
FET (Field Effect Transistor) compressors are fast and full of character. They add an aggressive snap to vocals, drums, and electric guitars. The 1176 hardware unit is the original FET compressor and one of the most-emulated pieces of gear in history.
The CLA-76 from Waves is the go-to plugin emulation. Reach for FET when you want compression you can hear, not the invisible kind.
Optical compressors (often called opto comps) use a light source and a photoresistor to detect the signal. The result is slow, smooth compression that sounds natural on vocals and bass. They're forgiving and don't punch transients the way FETs do.
The Teletronix LA-2A is the classic, with plugin versions from various developers. If a vocal sounds harsh on an 1176 but still needs leveling, the LA-2A almost always solves it.
Vari-Mu compressors use tube circuitry that changes how it compresses based on the input signal. They're slower, warmer, and more organic than VCA or FET designs. Great on mix buses and mastering chains where you want compression to feel like part of the music.
The Fairchild 670 is the original. The SPL IRON is a modern vari-mu-style compressor that's especially good for mastering. Shadow Hills sits in this territory too.
Digital compressors aren't based on any specific hardware. They can be transparent, colourful, surgical, or musical depending on how the developer designed them. The advantage is flexibility: visual feedback, sidechain options, oversampling, and features no analog unit could ever do.
FabFilter Pro-C 2 is the reference modern digital compressor and the one I keep coming back to. FabFilter Pro-MB is the go-to for multiband work. Digital compressors won't replace analog character when you need it, but they're the right choice when you want precision and visual control.
The Parallel Compression Trick (And When To Use It)
Parallel compression is one of the most useful tricks in modern mixing, and most beginners don't know it exists. The idea is simple: instead of compressing a track directly, you blend a heavily compressed copy of the signal underneath the original.
Why does it work? Heavy compression alone kills the dynamics and makes things sound flat. But a heavily compressed copy mixed underneath an uncompressed signal gives you the best of both worlds. You keep the natural transients and dynamics of the original, and you get the body and sustain of the compressed copy.
Where I use parallel compression most:
- Drum buses: Adds weight and body without making the drums feel squashed.
- Lead vocals: Brings up the breath, room tone, and quiet details without affecting the loud parts.
- Bass guitar: Adds sustain and density without losing the pick attack.
Starting settings on the parallel send: ratio around 6:1 to 10:1, fast attack, fast release, threshold low enough to get noticeable gain reduction (8-12 dB).
Then blend it under the dry signal until it's just barely audible. The trick is that you should be able to bypass the parallel send and clearly notice the mix gets smaller and thinner. If bypassing it sounds like nothing changed, the level is too low. If it sounds drastically different, the level is too high.
Sidechain Compression: How to Make a Kick and Bass Coexist
Sidechain compression is when a compressor reacts to one signal but applies the gain reduction to another. The most common example: a kick drum triggers a compressor on the bass, ducking the bass every time the kick hits.
Why this matters: kick and bass live in the same frequency range. When they hit at the same time, they fight for space and the low end turns muddy. Sidechain compression ducks the bass momentarily when the kick hits, so the kick punches through and the bass fills the space between hits.
This technique is everywhere in EDM, hip-hop, and modern pop. The pumping effect in house music is just very obvious sidechain compression with extreme settings.
Where I use sidechain compression:
- Bass ducking under kick: The classic move, especially for electronic music.
- Pad ducking under vocals: So the pad doesn't crowd the vocal.
- Reverb ducking under the dry signal: Keeps the reverb tail clear of the main transients.
Most modern compressors have a sidechain input. In FL Studio you can use the built-in Fruity Limiter in compressor mode (see image below), which makes sidechain routing especially easy. Just send the trigger signal to the sidechain input of the compressor and it'll react to that source instead of the main signal.
Dynamics Tools Compared: When To Use Each
A compressor isn't the only tool that controls dynamic range. Limiters, clippers, and expanders all do related jobs. Knowing the difference saves you from using the wrong tool.
| Tool | What it does | When to use it |
| Compressor | Reduces dynamic range gradually based on threshold and ratio | Tonal shaping, leveling, glue |
| Limiter | Hard ceiling that prevents the signal from going above a set level | Mastering, peak control, loudness |
| Clipper | Cuts off peaks abruptly by hard-clipping the waveform | Loudness, aggressive drums, character |
| Expander | Increases dynamic range by making quiet parts quieter | Noise reduction, gating, depth |
How to Use a Compressor: My 6-Step Process
This is the workflow I follow on most tracks. The order matters more than the exact settings, because the right settings depend on the source.
Start with a moderate ratio around 4:1. Aggressive settings come later
Load the compressor on the track. Set a starting ratio between 2:1 and 4:1 for most sources. Higher ratios sound more obvious and aggressive. Lower ratios feel transparent. Save extreme ratios (8:1 and above) for parallel compression or drum bus work where you want to hear the compression as an effect.
Aim for 3-5 dB of gain reduction on most sources. More than 6-8 dB is usually too much
Lower the threshold until the gain reduction meter starts moving. Aim for 3-5 dB of gain reduction on most sources. If you're seeing 10 dB+, you're either compressing too hard or the source has too much dynamic range to begin with. Big gain reduction numbers usually point to a problem with the source, not a problem the compressor should solve alone.
Slow attack to keep transients punchy, faster release for a more lively feel
This is where compression becomes an art. Slower attack times preserve transients (great for drums that need to punch). Faster attack times catch peaks (great for vocals that need leveling). Release controls how quickly the compressor lets go. Faster releases feel pumping and aggressive. Slower releases feel smooth and natural.
Match the compressed level to the dry level before deciding if the compression is helping
Compression reduces the overall level. Add makeup gain to bring it back to where it was before. Most modern plugins have an "auto gain" mode that does this for you. Manual control gives more precision. Critical: always match levels before you A/B. Louder always sounds better to the brain, so without level matching you'll think every compression move is an improvement.
Hit bypass. If the compressed version isn't a clear improvement at matched levels, undo it
Bypass the compressor and listen to the raw signal. Un-bypass and listen to the compressed version. If the compressed version isn't clearly better in the mix context, you don't need that compressor. Compression isn't a default move on every track.
Always evaluate compression in the context of the full mix, not in solo
A compressed track might sound great in solo but fight with other elements in the mix. Always take the solo off and listen to how the compressed track sits with everything else. Sometimes you'll add compression and discover that the track now masks the vocal or fights with the snare. That's information. Adjust or remove.
For a deep dive on the EQ side of the chain, including whether to EQ before or after compression, check our EQ guide.
I used to compress every track with the same settings. Same threshold, same ratio, same attack and release. The result was mixes that all sounded the same and never quite right.
Once I started setting compression by source (FET-style fast attack on snares, slow attack on kicks, optical-style slow attack on vocals) my mixes started having actual depth. There's no universal "good compression setting." Every source needs its own approach.
Compressor Settings Cheatsheet by Source
Pick a source below to see the compressor settings I'd reach for as a starting point, what the goal of the compression is, and what to listen for.
The vocal should sound more consistent without sounding squashed. If you can hear the compressor "pumping" on consonants like S and T, your attack is too fast. If the loud lines still jump out, lower the threshold or raise the ratio.
These are starting points, not rules. A pop vocal recorded on an SM7B needs different compression than an EDM lead vocal recorded through a Neumann into a clean preamp. Use these to know where to begin. Use your ears to make the actual call.
Compression Tips That Will Save You Hours
After enough sessions, certain habits become reflex. These aren't rules. They're shortcuts that work in most situations.
- Aim for 3-5 dB of gain reduction: More than that and you're usually fighting the source instead of shaping it. If you need 10+ dB of gain reduction to control a track, the recording is probably the real problem.
- Gain-match before you A/B: Louder always sounds better. Without level matching, every compression move feels like an improvement, even the bad ones. Most modern compressors have an auto gain-match or output level control. Use it.
- Use the gain reduction meter, not the output meter: The output meter tells you the level. The gain reduction meter tells you what the compressor is actually doing. Watch that one to make decisions.
- Don't compress on the way in: Some old-school engineers compress during recording. With modern DAWs and plenty of headroom, that's not necessary. Compressing on the way in is permanent. Compressing in the mix is reversible.
- Two light compressors beat one heavy one: If you need a lot of compression, split it across two compressors with different settings instead of cranking one. Sounds more natural and gives more control over how the compression behaves.
- Save your favourite chains as presets: When you find a vocal chain that works for your style, save the whole chain (compressor settings, gain, even the plugin order). It's a starting point for the next song, not a final solution.
- Listen with fresh ears: Compression decisions made after an hour of mixing are usually wrong. Step away. Come back. If the compressor still feels right, it's right. If it now feels like too much, it was too much.
Common Compressor Mistakes
The tips above will get you most of the way there. But there are a few specific traps that catch even experienced producers, usually because they don't feel like mistakes while you're making them.
- Compressing everything: Not every track needs compression. A vocal recorded by a good singer in a controlled environment might not need any. A perfectly programmed kick sample doesn't need compression. Use compression when it solves a problem or shapes a sound, not as a default move.
- Squashing the life out of a track: Aggressive compression flattens dynamics. Sometimes that's what you want (parallel compression, drum bus glue). Most of the time, you want to keep some natural movement. If a track sounds dead and lifeless after compression, you're over-compressing.
- Using the same settings on every source: A drum bus and a vocal need totally different compression. A bass and a snare need totally different compression. Default presets are starting points, not solutions. Adjust by ear for every source.
- Ignoring the attack time: Fast attack on a snare kills the crack. Slow attack on a vocal lets the loud peaks slip through. Attack time is the parameter that most controls what compression sounds like, but it's the one beginners pay the least attention to.
- Not matching gain when comparing: Compressed signal has a lower output level. Compressed signal with makeup gain is louder than the dry signal. Both make A/B comparisons impossible unless you match levels first. Most plugins have an auto gain-match feature. Turn it on.
- Using compression to fix bad recordings: A bad take with too much variation can't be saved with compression. A poorly mic'd drum kit won't sound right no matter what compressor you reach for. Compression shapes what's there. It can't fix what's missing.
- Setting threshold and ratio without listening to the gain reduction meter: Two tracks at the same output level can need totally different threshold settings if their dynamic range is different. Always look at the gain reduction meter and adjust from there, not from preset numbers.
Final Thoughts
Compression takes longer to learn than EQ. EQ is mostly about hearing what's wrong and removing it. Compression is about feel, character, and intent, which all require time at the controls to develop.
The fastest way to improve is to compress more deliberately. Don't just turn the knob until it sounds different. Set the threshold first, then choose the ratio, then dial attack and release by ear. Listen for what changes. After a few months of conscious compression practice, your hands will start reaching for the right settings before you've thought about them.
Once the fundamentals click, the right tool makes everything faster. The plugins on our best compressor plugins guide are the ones I keep reaching for after years of mixing.
FAQs
How does a compressor work?
A compressor monitors the input signal level and applies gain reduction whenever the level exceeds a set threshold. The amount of gain reduction is controlled by the ratio. The speed of the reaction is controlled by attack and release. The makeup gain brings the output level back up to compensate for the volume reduction caused by compression.
Should I EQ before or after a compressor?
Both, depending on the goal. Corrective EQ before a compressor removes problem frequencies so the compressor doesn't react to them. Tonal EQ after a compressor shapes the final sound. My typical chain on a vocal is corrective EQ first, then compressor, then a final tonal EQ.
What's the difference between a compressor and a limiter?
A limiter is essentially a compressor with an extreme ratio (usually 20:1 or higher) and a brick-wall ceiling. Compressors gradually reduce gain based on threshold and ratio. Limiters prevent the signal from going above a specific level no matter what. Use a compressor for shaping and a limiter for peak control or final loudness.
How much gain reduction is too much?
For most sources, 3-5 dB is the sweet spot. Anything beyond 6-8 dB starts to sound obviously compressed. If you need 10+ dB of gain reduction to control a track, the source has too much dynamic range to begin with, and you should probably split the work between two compressors or fix the source.
What is the difference between VCA, FET, and Optical compressors?
VCA compressors are fast and clean, great for bus work. FET compressors are fast and aggressive, great for drums and vocals that need character. Optical compressors are slow and smooth, great for vocals and bass that need natural levelling. Each has a different sound and reacts differently to the same input signal.
What is parallel compression?
Parallel compression is when you blend a heavily compressed copy of a signal underneath the dry original. You get the body and sustain of heavy compression while keeping the natural dynamics of the original. Common on drum buses and lead vocals.
What is sidechain compression?
Sidechain compression is when a compressor reacts to one signal but applies the gain reduction to another. The most common example is a kick triggering a compressor on the bass, ducking the bass every time the kick hits. Used heavily in EDM and modern pop for the pumping effect.
How do I use a multiband compressor?
A multiband compressor splits the signal into frequency bands and compresses each band independently. Use it when you need to control a specific frequency range without affecting the rest of the signal, like taming sibilance on a vocal (high-band compression only) or controlling a boomy low end on a mix (low-band only). It's surgical, so use it when a regular full-range compressor would over-compress.
Can I use compression on every track?
You can but you shouldn't. Some tracks benefit from compression, some don't need it, and some are made worse by it. Compress when there's a specific reason: levelling a dynamic performance, adding punch, shaping tone, gluing a group of sounds. Don't compress as a default.
Why does my mix sound worse after I add a compressor?
Usually one of three things. You're compressing too hard (lower the threshold or the ratio). You forgot to gain-match (use the auto makeup or output level). Or the track didn't need compression in the first place (bypass and see if the dry version sounds better).
