Audio compression reduces the dynamic range of a sound by turning down its louder moments. In music production, a compressor can make a vocal more consistent, control uneven bass notes, change the punch of a kick drum or create the rhythmic pumping heard in dance music. It does not automatically improve every track, and it is not the same thing as reducing the size of an MP3 or WAV file.
To use compression well, you need to understand six controls: threshold, ratio, attack, release, knee and makeup gain. The real skill is deciding whether a sound needs compression at all, then adjusting those controls while listening rather than following a preset blindly.
This guide explains what compressors do, shows practical settings as starting points, and walks through compression for techno, house and other electronic music. If you’re learning production from the beginning, start with our introduction to digital audio workstations.
What is audio compression in music?
Every recording has a dynamic range: the difference between its quieter and louder parts. A kick sample has a strong initial transient followed by a tail. A vocal performance has words and syllables at different levels. A live bass recording may include notes that jump forward more than others.
A compressor monitors the incoming signal. When its detected level goes above a chosen threshold, the compressor reduces the gain according to its ratio and timing behaviour. Once the signal falls, its gain reduction recovers. The quieter moments are not automatically boosted by the main compression action. They may become relatively more audible when you restore overall level with makeup gain.
Compression is therefore both a technical and creative tool. You can use it subtly to keep a performance even, or deliberately to alter transients and make a groove breathe. The right outcome depends on the source and your musical intention.
Audio compression versus file compression
The phrase “audio compression” describes two different processes:
| Type | What changes? | Example |
|---|---|---|
| Dynamic range compression | How a signal’s level varies over time | A compressor plugin reducing loud vocal peaks |
| Data or file compression | How audio information is encoded and stored | Encoding a WAV as a smaller MP3 or AAC file |
This article covers dynamic range compression. A plugin compressor does not normally shrink your project file or make a recording take up less storage.
How does an audio compressor work?
A compressor has a signal detector, a gain-control process and timing behaviour. At a simplified level, it measures the incoming sound, decides how far the level exceeds a threshold, and reduces gain accordingly. In real plugins, the detector may use peak or averaged measurements, soft knees, program-dependent timing and other design choices.
Imagine a compressor with a threshold of −18 dB and a 4:1 ratio. A sustained signal that reaches −10 dB is 8 dB above the threshold. With an ideal hard-knee compressor, once its gain-reduction envelope has fully settled, only 2 dB of that excess remains at the output. The level becomes approximately −16 dB, a reduction of 6 dB before makeup gain.
| Input level | Over −18 dB threshold | Ideal output at 4:1 | Gain reduction |
|---|---|---|---|
| −22 dB | Below threshold | −22 dB | 0 dB |
| −18 dB | At threshold | −18 dB | 0 dB |
| −14 dB | 4 dB | −17 dB | 3 dB |
| −10 dB | 8 dB | −16 dB | 6 dB |
These are ideal steady-state figures, not a prediction of every transient. Attack, release, detector design, knee and plugin character can produce a different moment-by-moment result.
A common misunderstanding is that a compressor “makes loud sounds quiet and quiet sounds loud” in one step. It usually reduces loud passages first. Raising the resulting output with makeup gain is a separate action, and it raises quieter material and noise as well.
Compressor settings explained: the six essential controls
1. Threshold: when compression starts
The threshold sets the level at which the detector begins applying gain reduction, normally measured in dB. Lowering it makes more of the signal eligible for compression. Raising it restricts the effect to louder sections.
Instead of copying a threshold value from a tutorial, watch your compressor’s gain-reduction meter. A kick sample peaking at −8 dB and a bassline peaking at −20 dB will not respond the same way to a threshold of −16 dB.
2. Ratio: how strongly the compressor reduces level
The ratio describes the relationship between increases above the threshold at the input and output. At 2:1, 2 dB above threshold becomes about 1 dB above threshold once the compressor settles. At 4:1, 4 dB becomes about 1 dB. At 1:1, an ordinary downward compressor applies no compression.
A low ratio can be useful for gentle levelling. A higher ratio increases control or creates a stronger effect, but it is not automatically better for drums, vocals or bass.
3. Attack: how quickly gain reduction develops
Attack describes how quickly the compressor responds after the signal exceeds the threshold. With a very fast attack, the initial hit of a drum can be reduced; with a slower attack, more of the transient may pass before the gain reduction takes hold.
This is why setting a fast attack does not always create a punchier kick. If your goal is more impact, a slower attack can preserve the initial click while controlling the body. If your goal is to tame a sharp, aggressive transient, a faster attack may help.
Numbers are not universal: compressor designs can interpret the same displayed attack time differently. Judge the sound and the meter rather than the millisecond value alone.
4. Release: how the compressor recovers
Release sets how quickly gain reduction returns toward zero after the detector no longer calls for it. A release that is too fast for a bass signal can cause audible distortion or unwanted modulation. A release that is too slow can leave the compressor reducing the next hit or musical phrase.
In electronic music, release is closely connected to groove. You may want a bassline to recover between kicks; for a pumping effect, you may deliberately make the return of the level part of the rhythm. Auto-release modes can be a useful starting point, but listen for unnatural surging or lifeless dynamics.
5. Knee: how gradually compression begins
A hard knee shifts more abruptly into the selected ratio around the threshold. A soft knee introduces gain reduction gradually over a range around it. This can make compression less conspicuous on vocals, pads and buses. It is not a quality ranking: a hard knee can be exactly what a rhythmic effect requires.
6. Makeup gain and output level
Compression turns down parts of your signal. Makeup gain allows you to restore the resulting overall level. Be careful: a processed version that is simply louder often seems better even when its tone and dynamics are worse.
Use bypass comparisons at approximately the same perceived loudness. Some compressors offer automatic makeup gain, but it is a convenient estimate, not proof that two versions are level-matched.
Quick compressor settings reference
| Control | When you increase it | What to listen for |
|---|---|---|
| Threshold (raise) | Fewer moments cross it | Less gain reduction |
| Ratio | More reduction above threshold | Control versus flattened dynamics |
| Attack time | Compressor reacts more slowly | More or less initial transient |
| Release time | Gain reduction recovers more slowly | Groove, pumping or loss of movement |
| Soft knee | Onset is more gradual | Smoother, less obvious compression |
| Makeup gain | Processed output becomes louder | Unwanted loudness bias or clipping |
How to use a compressor: a seven-step beginner workflow
This procedure works with most stock DAW compressors, including ones in Ableton Live, FL Studio, Logic Pro, Cubase and REAPER. The exact knob positions depend on the source, so the gain-reduction meter and your ears are the reference points.
- Identify the problem. Is the bass uneven? Are vocal peaks distracting? Do you want drum punch or deliberate rhythmic pumping? If you cannot describe the problem, leave the compressor bypassed for now.
- Place a compressor on the relevant track or bus. Start with a stock compressor that displays gain reduction clearly.
- Begin with a moderate ratio. Try 2:1 or 3:1 as a learning starting point, then adjust to taste. It is not a mandatory rule.
- Lower the threshold until the loud moments trigger reduction. For gentle control, listen while aiming for a few dB of movement rather than compressing continuously without a reason.
- Adjust attack to shape the transient. Move between faster and slower settings while listening to the initial hit or consonants in a vocal.
- Adjust release to follow the phrasing or beat. Watch the meter recover between events. If the mix breathes awkwardly, slow it down or speed it up as appropriate.
- Level-match and bypass. Use makeup gain only as needed, toggle the plugin and decide whether the compression actually helps in the full mix.
Save a version of your project before making aggressive changes. Sometimes lowering the clip gain, editing note velocities or adjusting volume automation solves the problem with less processing.
Practical audio compression examples for electronic music
Kick drum: control versus punch
With a kick drum, first decide whether you want to reduce the initial click, smooth an uneven live performance or emphasize impact. Many electronic kicks are already carefully designed and heavily processed, so another compressor may do very little beyond removing transients.
Experiment: On a kick that has a sharp transient and a long body, try a ratio around 3:1 or 4:1, begin with an attack near 15–30 ms and a release that recovers before the next beat. Lower the threshold to hear a few dB of reduction. Then shorten the attack dramatically and compare. You may find the slower attack lets the initial hit through while the faster attack softens it. Use the version that supports the groove.
Those values are illustrative, not recommended defaults for every kick. A short sampled 909-style kick, an acoustic drum recording and a distorted industrial kick will react very differently.
Bass: keep notes even without flattening movement
A bassline with inconsistent note levels can benefit from modest compression, particularly when a filter resonance or different notes create uneven peaks. Try gentle ratios such as 2:1–4:1, adjust the threshold based on actual gain reduction, and choose attack and release times that preserve the musical envelope.
If low notes start buzzing or distortion appears, experiment with a longer release, less gain reduction or a different detector setting. Compressing everything harder may make low-end problems worse. For deeper kick-and-bass arrangement and EQ techniques, see our low-end mixing guide.
Drum bus: help separate hits feel cohesive
A drum bus combines several percussion tracks. Gentle compression can shape their collective movement, but it can also let a loud kick trigger reduction that drags down hi-hats and claps. Begin conservatively, check whether the compressor responds mainly to the kick, and consider detector filtering or parallel compression if needed.
Listen to the snare’s attack and the hi-hat’s decay. If the percussion becomes dull, try a slower attack, less gain reduction or no compression at all. If it starts audibly pumping, decide whether that supports the music or distracts from it.
Vocals and pads: manage level changes without obvious pumping
For uneven vocals, use reasonable clip gain or volume automation for big differences first, then compression to smooth shorter variations. A moderate ratio, a knee that engages gradually and attack/release settings that follow the performance are a sensible starting approach. Don’t rely on compression to fix poor microphone placement or clipping already present in the recording.
Pads may have little need for compression unless they contain peaks, awkward envelope changes or a deliberate rhythmic effect. A sustained sound with stable levels can become flatter for no practical benefit.
Sidechain compression: make room for the kick
Sidechain compression uses a separate control signal to decide when the compressor acts. In dance music, the familiar use is to lower the bass level briefly whenever the kick plays. That can improve rhythmic clarity or create an obvious pumping sound.
Try this in a DAW that supports external sidechain inputs:
- Place a compressor on the bass track, not the kick.
- Enable its external sidechain input and select the kick track as the trigger.
- Set a moderate or strong ratio for an audible demonstration and lower the threshold until each kick causes bass gain reduction.
- Use a fast-enough attack to make space at the kick onset, then set the release so the bass recovers musically between hits.
- Reduce the intensity until the kick and bass sit together. Compare with bypass to check that you have improved separation rather than simply made the mix quieter.
Not every electronic track needs audible ducking. For a techno groove, you might want barely noticeable movement; for house or other dance styles, the pumping itself may be part of the arrangement. A volume-shaper or automation can also create deliberate ducking, but that is not detector-driven compressor behaviour.
There is also sidechain filtering, where you change what the compressor’s detector hears without necessarily changing the audible tone of the track. For example, high-passing the detector can stop deep bass from causing excessive whole-mix gain reduction.
Parallel compression, limiting and multiband compression
Parallel compression
Parallel compression mixes a compressed version with an uncompressed version. This lets you introduce density or sustain while keeping some of the original transient structure. Many plugins include a dry/wet control; alternatively, you can use a parallel return or duplicated signal path. Watch for differences in timing and latency if your setup doesn’t compensate automatically.
Limiting and lookahead
A limiter is a dynamics processor designed to restrict peaks more firmly than ordinary gentle compression. Some lookahead limiters delay the audio to anticipate sudden events. They can help control the final output, but no limiter can retroactively repair distortion that was already recorded or caused by overloaded processing earlier in the signal path. Pushing a limiter too hard can create pumping, distortion and fatigue.
Multiband compression and dynamic EQ
A multiband compressor splits audio into frequency regions and compresses those regions independently. This can be useful when, for example, low-frequency energy varies more than the rest of a mix. A dynamic EQ typically adjusts specific frequency bands when the detected signal reaches a condition, sometimes with narrower, more surgical control.
These tools are more complex than ordinary broadband compression and can introduce trade-offs such as crossover artefacts or unexpected tonal shifts. Reach for them when you hear a specific frequency-dependent problem. Our broader dynamic processing guide explores these ideas further.
Peak versus RMS detection: why two compressors can sound different
A peak detector follows short level spikes more readily, while an RMS-style detector responds to an averaged signal level and may be less sensitive to extremely brief transients. Neither is automatically superior. Compressor character, envelope design and attack/release behaviour also matter.
This is why two compressors displaying identical threshold, ratio and timing values can still sound different. Use the numbers as a guide, not as proof that the processing is equivalent. Ableton’s Compressor documentation provides a useful explanation of its Peak and RMS modes.
Five common audio compression mistakes
- Compressing everything by default. Start with a reason. A consistent sample may need no further dynamics control.
- Assuming fast attack always equals punch. Fast attack often reduces the transient that gives a hit its impact.
- Setting release without listening to the tempo and signal. Recovery that overlaps the next hit can change the groove or cause unwanted pumping.
- Comparing different output levels. Match perceived loudness when bypassing a compressor, or a louder processed version may seem better for the wrong reason.
- Confusing compression with repair. Compression will not remove room echo, distortion already baked into a recording, masking caused by poor arrangement or phase cancellation.
Do you need an expensive compressor plugin?
No. Stock DAW compressors are capable of teaching every major concept in this article and often offer sidechain routing, gain-reduction meters and different detection modes. Start there. A third-party plugin may provide an interface or response you prefer, but the basic questions remain the same: what triggers gain reduction, how much, how fast, and does the result sound better?
If you want to study a detailed commercial design, the official FabFilter Pro-C 2 manual documents threshold, knee, range, detector options, lookahead, sidechains and parallel controls. You do not need to own it to learn from the explanations. For other production fundamentals, see our guide to mixing electronic music.
Audio compression FAQ
What is the purpose of audio compression?
Its main purpose is to control dynamic range, making loud moments less dominant. It can also shape the attack and sustain of a sound, even out a performance, or create rhythmic effects.
Does compression make music louder?
Compression itself normally reduces loud passages. The resulting headroom can allow makeup gain to raise the overall level, increasing perceived loudness. This is not the same as saying every compressed sound becomes louder automatically.
What ratio should I use for a vocal, kick or bass?
There is no fixed ratio for a particular instrument. A gentle 2:1 or 3:1 can be a useful learning starting point; more assertive processing may call for something stronger. The source signal, threshold, attack and release are as important as the ratio.
Should I compress every kick drum?
No. Many produced kick samples have already been compressed, clipped, limited or otherwise shaped. Compress only if you need to change the envelope, control actual variation or create a specific effect.
What is the difference between compression and limiting?
Ordinary compression commonly uses moderate ratios for dynamics control. Limiting generally applies stronger peak restriction, often with high ratios and specialized timing. A limiter can help manage output peaks, but extreme limiting can audibly change the sound.
What is sidechain compression in techno?
It is a method where a trigger, often the kick drum, causes the compressor on another track such as bass to reduce gain. This can make space at each kick or create a rhythmic pulse.
When should I use parallel compression?
Try it when heavy compression gives useful body or sustain but removes too much of the original punch. Blend a compressed path with the dry sound and level-match the result.
Can compression fix clipping or a distorted recording?
No. A compressor cannot recover audio information that has already been clipped or distorted earlier in the recording chain. Prevent overloads at the source, and use dynamics processing as a creative or level-control tool rather than a universal repair.
Final thoughts: let the music decide
A compressor is one of the most useful tools in electronic music, but its real value comes from understanding the relationship between level, timing and musical movement. Threshold and ratio determine which parts are turned down and by how much; attack and release determine how the envelope changes through time. A few dB of well-judged gain reduction may help more than aggressive processing on every channel.
Make a simple loop, load your DAW’s compressor, and practice hearing those differences. Keep bypassing the effect and asking whether the groove genuinely improves. Then use the techniques where the music calls for them.
Sources and further reading
- Ableton Live 12 Manual: Compressor, Glue Compressor and Multiband Dynamics (threshold, ratio, envelope, detection and sidechain behaviour)
- FabFilter Pro-C 2 official manual (knee, lookahead, release models, makeup gain and parallel compression)
- iZotope: What Is Audio Compression? (signal behaviour, practical listening methods and dynamic smoothing)
- Sound On Sound: Compression Made Easy (transients, envelope timing and audible distortion)
