A studio compressor unit with rows of knobs beside coiled cables and headphones on a warmly lit desk.

Compression is one of those words that shows up everywhere in music production. People say compress the vocal or add some glue on the mix as if it is obvious. Then you load a compressor and see threshold, ratio, attack, release, knee, makeup gain, sidechain, and meters bouncing in a way that feels like a cockpit. At its core, though, compression is simple: it turns down loud parts so the overall signal feels more controlled and consistent. The art is in deciding what loud means and how quickly the compressor reacts.

What is compression in audio?

Compression is dynamic range control. It reduces the difference between the loudest and quietest parts of a signal. When a vocalist hits a loud note, a compressor can gently reduce that peak so the vocal stays in place. When a bass note jumps out, compression can even it out. On drums, it can reshape the transient and the sustain. None of this is automatically good. Too much compression can make a track lifeless, harsh, or small, which is why intent matters more than the amount.

The key controls in plain language

Most compressors share the same core controls. The threshold is the level where compression starts: below it, nothing happens; above it, the compressor begins reducing gain. The ratio controls how strongly it turns down anything above the threshold, from a gentle 2 to 1, to a moderate 4 to 1, to an aggressive 8 to 1 and beyond. A limiter is essentially an extreme ratio.

Attack is how quickly the compressor reacts after the signal crosses the threshold. A fast attack tames peaks but can remove punch, while a slower attack lets the transient through and keeps things alive. Release is how quickly it stops compressing once the signal falls back below the threshold. Too fast causes pumping; too slow clamps down and never recovers between notes. Finally, because compression lowers the signal, you add makeup gain to bring the level back up. This is where beginners get tricked, since louder often sounds better, so always level-match when comparing.

What does compression sound like?

Compression can do several different jobs. It can control peaks so a track is smoother and less spiky. It can add sustain so notes feel longer. It can add density so the track feels closer and more present. It can change punch, making drums feel tighter or softer depending on your attack. And it can glue a bus so a group of tracks feels unified. If you are not sure what to listen for, watch the gain reduction meter and toggle bypass while keeping levels matched.

Starting settings for common sources

There is no universal preset, but these are safe starting points. For vocals, aim to control peaks without crushing: a ratio of 2 to 1 up to 4 to 1, a medium attack that lets consonants through, a medium release timed to the phrase, and only a few decibels of gain reduction on loud words. For bass, aim for consistent low end with a ratio around 3 to 1 up to 6 to 1, a medium-fast attack, a medium release, and steady gain reduction on each note.

For individual drums, remember that a faster attack reduces the transient while a slower attack keeps the punch; start around 4 to 1 and adjust by ear. On a drum bus, try a slower attack, a medium release, and only a little gain reduction for cohesion. On the mix bus, be gentle if you do it at all: a low ratio around 1.5 to 1 or 2 to 1, a slow attack, a medium release, and just one to two decibels of gain reduction. If you hear obvious pumping, back off.

Common mistakes

Compression mistakes are predictable. The most common is over-compressing simply because louder feels better in the moment. Others include using a fast attack everywhere and losing all your punch, setting the release too slow so the groove flattens, and stacking compressors with no clear reason. If you do not know what problem you are solving, bypass the compressor and fix the arrangement or the performance first. Compression cannot rescue a part that is not working.

Bottom line

Compression is a tool for controlling dynamics. Learn threshold and ratio first, then use attack and release to shape the feel. Start gentle, always level-match, and compress with a clear goal in mind: control peaks, add sustain, or glue a bus. Used with intent, it makes a mix feel steady and finished. Used carelessly, it flattens everything that made the performance interesting.

For Sunday readers

Subscribe to the Sunday Stem

A short, honest dispatch on American music, three mornings a week, with the Sunday Stem on craft, catalog, and the writers keeping the long tradition alive.

More from the Song Production desk →

Frequently asked

What is compression in audio, simply put?

Compression is dynamic range control. It reduces the difference between the loudest and quietest parts of a signal. When a vocalist hits a loud note, a compressor gently turns that peak down so the vocal stays in place; when a bass note jumps, compression evens it out. It is not automatically good, though. Too much compression can make a track feel lifeless, harsh, or small, so it is a tool to use with intent rather than by default.

What do threshold and ratio actually do?

The threshold is the level where compression starts. Below it, nothing happens; above it, the compressor begins reducing gain. The ratio controls how strongly it turns down the signal that crosses the threshold. Around 2 to 1 is gentle, 4 to 1 is moderate, and 8 to 1 or higher gets aggressive. A limiter is essentially compression at an extreme ratio. Together, threshold and ratio decide how much gain reduction you are applying.

How do attack and release affect the sound?

Attack is how quickly the compressor reacts once the signal crosses the threshold. A fast attack tames peaks but can remove punch, while a slower attack lets the transient through, keeping drums and vocals lively. Release is how quickly it stops compressing after the signal drops back down. Too fast can cause audible pumping; too slow can clamp the sound and never recover between notes. These two controls shape feel and groove more than most beginners realize.

What are safe starting settings for vocals?

For vocals, the goal is to control peaks without sounding crushed. A ratio around 2 to 1 up to 4 to 1 is a good start, with a medium attack so consonants come through and a medium release timed to the phrasing. Aim for just a few decibels of gain reduction on the loudest words. Then level-match and bypass to check that it actually sounds better, not just louder. Adjust by ear from there rather than chasing exact numbers.

Further reading on From The Stem

· What Is Gain Staging?
· What Is Reverb?
· What Is a Vocal Comp?