Compression reduces the volume difference between the loudest and quietest parts of an audio signal. It makes quiet moments more audible and prevents loud peaks from overwhelming the mix. In practical terms, compression is what makes a vocal sit consistently on top of a beat instead of disappearing in the verse and spiking in the chorus.
Compression is the most misunderstood tool in production. Not because the concept is hard, but because most explanations bury the practical value under technical jargon. You do not need to understand the math behind a VCA circuit to use a compressor well. You need to understand what it does to sound and when to reach for it. This article covers both. For the full production workflow including where compression fits, see Music Production Basics.
What Compression Actually Does to Sound
Imagine a singer who whispers the verse and belts the chorus. Without compression, you have two choices: set the level for the verse (and the chorus clips) or set the level for the chorus (and the verse disappears). Compression solves this by turning down the loud parts automatically, narrowing the gap between soft and loud.
The result is a signal that sits at a more consistent level. After compressing, you can raise the overall volume of the track, which brings the quiet parts up without the loud parts going over the limit. That is the core value. Consistency.
The Five Controls That Matter
Every compressor, hardware or software, has the same fundamental parameters. Learn these five and you understand compression on any compressor you will ever use.
Parameter | What It Does | Starting Point |
|---|---|---|
Threshold | The volume level where compression kicks in | Set so the compressor activates on the louder phrases, not on everything |
Ratio | How much the signal is reduced once it crosses the threshold | 3:1 for vocals, 4:1 for drums, 2:1 for gentle glue |
Attack | How quickly the compressor reacts after the signal crosses the threshold | 10-30ms for vocals, 1-5ms to tame drum transients |
Release | How quickly the compressor stops compressing after the signal drops below the threshold | 50-100ms for most sources, faster for rhythmic pumping |
Makeup gain | Raises the overall level after compression | Match the compressed signal to the original perceived loudness |
Threshold is your on/off switch. Lower the threshold and the compressor works harder. Raise it and only the loudest peaks get compressed. Start with the threshold high and bring it down until you hear the compressor engaging on the parts that need control.
Ratio determines how aggressively the compressor reduces volume. At 2:1, a signal that exceeds the threshold by 10 dB gets reduced to 5 dB over. At 10:1, that same 10 dB overshoot becomes 1 dB. Ratios above 10:1 approach limiting, where the signal essentially cannot exceed the threshold.
Attack is where most of the character lives. A fast attack catches transients immediately, which softens the punch of drums or the consonants of a vocal. A slow attack lets the initial transient through untouched, preserving the snap and then compressing the sustain. On drums, slow attack keeps the hit punchy. On vocals, medium attack (10-20ms) controls volume without flattening the natural dynamics.
Release sets how quickly the compressor lets go. Too fast and the compressor pumps audibly between words or hits. Too slow and it stays clamped down, squashing everything after the first loud phrase. Match the release to the tempo and rhythm of the source.
Compression on Vocals
Vocals are the most common target for compression because the human voice has enormous dynamic range. A whispered line might be 20 dB quieter than a belted note. That range makes mixing impossible without some form of level control.
A typical vocal compression approach: set the ratio to 3:1 or 4:1, lower the threshold until you see 3-6 dB of gain reduction on the loudest phrases, use a medium attack (10-15ms) so the consonants still cut through, and set the release fast enough to recover between phrases but slow enough to avoid pumping. Then add makeup gain to bring the compressed vocal back to a natural level.
Many engineers use two compressors in series on a vocal. The first catches the biggest peaks with gentle settings. The second applies more consistent control. This avoids any single compressor working too hard, which sounds more transparent than one compressor doing all the work.
Compression on Drums
Drums use compression differently than vocals. The goal is usually to add punch, sustain, or both.
Kick and snare. A slow attack (20-30ms) lets the transient punch through, then the compressor grabs the body and sustain. This makes the hit feel powerful while controlling the ring. Ratio of 4:1 is a solid starting point.
Overheads and room mics. Heavy compression on room mics is a classic technique for aggressive, roomy drum sounds. Think of the big, explosive drum sound on records across rock and hip-hop. That sound is often a heavily compressed room mic blended with close mics.
Parallel compression. Blend a heavily compressed copy of the drums with the original uncompressed signal. This adds density and sustain without losing the natural dynamics of the performance. Most DAWs let you set this up with a send or a wet/dry mix knob on the compressor.
Bus Compression and Mix Glue
Applying gentle compression to an entire mix (or a group of instruments routed to a bus) is called bus compression. The goal is cohesion. A light ratio (1.5:1 to 2:1), a slow attack, and 1-2 dB of gain reduction on the peaks makes the mix feel like one performance instead of a collection of separate tracks.
Bus compression is where less is more. If you hear it working, you have gone too far. It should feel like the mix tightened up without any individual element changing noticeably. For a broader view of how compression fits into the mixing process, that guide covers levels, panning, EQ, and effects alongside compression.
If you are an independent artist mixing your own tracks, bus compression is often the difference between a mix that sounds like a collection of parts and one that sounds like a record. The other move that belongs to the mix rather than to any single track is sidechain compression, covered in the next section.
What Is Sidechain Compression
Every compressor watches a signal to decide when to work. Usually it watches the same signal it is compressing. Sidechain compression breaks that link: the compressor still processes one track, but its detector listens to a different track routed into a sidechain or key input.
The audible result is ducking. Each time the trigger signal crosses the threshold, the target signal drops, then comes back as the compressor releases. Nothing from the trigger track reaches the output, since its only job is to decide when the gain reduction happens. Line that up with a beat and the drop turns rhythmic, which is the pumping you hear on house, EDM, and a lot of pop.
The standard use case is kick and bass. Both sit in the same low frequencies, so when they land together the low end turns to mush and neither reads clearly. Turning the bass down for the whole track fixes the collision and costs you weight everywhere else. Sidechaining the bass to the kick moves it out of the way for the length of the hit and leaves it alone the rest of the time.
A starting point on kick and bass. Put the compressor on the bass (or on the 808, if a separate kick sample sits on top of it) and feed the kick into its sidechain input. Set the ratio to 4:1, then bring the threshold down until each kick pulls 3 to 6 dB of gain reduction off the bass. A fast attack (1-5ms) ducks the bass shortly after the kick triggers the compressor, and a slower attack lets more of the bass through before full gain reduction.
Release is the control tied to tempo. At 120 BPM a beat lasts 500ms, so a kick on every beat gives the bass that long to recover. Start in the 50-100ms range and lengthen it until the bass swells back instead of snapping back. A sparser kick pattern tolerates a longer release, while a busy one needs a shorter release or the bass never fully returns.
The same routing works anywhere one part needs to get out of another's way. A pad under a lead vocal, a guitar bed under a rap verse, a reverb return that ducks under the vocal feeding it. The trigger changes. The five controls do not.
When Not to Compress
Compression is not mandatory on every track. If a vocal performance has consistent dynamics, adding compression just because "vocals always need compression" can kill the natural feel. If a guitar part already sits well in the mix, compressing it adds nothing.
The rule: compress to solve a problem, not to check a box. If the problem is inconsistent volume, compress. If the problem is lack of punch, compress. If there is no problem, leave the fader where it is. For the EQ side of the equation, the same principle applies: fix problems before adding character.
Frequently Asked Questions
What is a good compression ratio for vocals?
Start at 3:1 for general vocal compression. Use 4:1 or higher for more aggressive control on dynamic performances. Lower ratios like 2:1 work for gentle leveling on already-consistent takes.
What is the difference between compression and limiting?
A limiter is a compressor with a very high ratio (typically 10:1 or higher). It prevents the signal from exceeding a set ceiling. Limiters are used in mastering to set the final loudness.
Does compression make audio louder?
Not directly. Compression reduces the loud parts. Makeup gain then raises the overall level. The result is a louder average level because the peaks are no longer eating up the headroom.
Can you over-compress?
Yes. Over-compression squashes dynamics, makes tracks sound lifeless, and introduces audible pumping or distortion. If the mix feels flat and airless, reduce the ratio or raise the threshold.
Read Next:
Production Gets Complicated. Your System Should Not.
The mix is yours to get right. What comes after it is a different job: release dates, campaign timing, and keeping every project and track in one place. Orphiq handles that side, and Apollo, Orphiq's AI music strategist, generates a release timeline working backward from your release date when you ask for one.