Updated October 2026 · Audio basics for beginners
A compressor reduces the gap between the loudest and quietest parts of a sound. When the signal rises above a threshold, the compressor turns it down by a set ratio. On speech it can even out the level and keep quiet words audible, but used too hard it makes the voice sound flat and lifeless.
Compression is one of the most common tools in audio, and one of the most misunderstood. Beginners often use it as a way to make everything louder, and then find that the voice sounds flat, tiring or squashed. A compressor is better understood as a tool for controlling the gap between loud and quiet, which can make speech easier to hear, but only when it is used with restraint.
This post explains what a compressor does, the main controls, the common settings for speech and the mistakes to avoid. It is a beginner guide, and the principles apply across many compressors.
A compressor watches the level of the signal. When the level rises above a set threshold, the compressor reduces the gain by an amount set by the ratio. The louder the signal above the threshold, the more it is reduced. The result is that loud parts are turned down, and the overall range becomes narrower. Quiet parts are not changed, so they can be raised later with gain.
Think of it as a hand on a volume knob that gently pulls down the loudest moments, so the rest can be turned up without the peaks clipping.
Speech has large differences between loud and quiet words, and listeners may need to adjust the volume to hear the quiet ones. A gentle compressor can reduce those differences so the voice is easier to follow on a phone or in a car. The effect is strongest on speakers who vary their volume a lot.
The aim is a natural, steady voice, not a squashed one. A small amount of compression often does more good than a large amount.
Compression can make a file sound louder, but it also reduces the dynamic range, which can make it feel tired. A recording that is compressed heavily to hit a loudness target may sound worse than one that is lightly compressed and then levelled. Use compression for the shape of the voice, and check the final loudness separately.
If a recording already sounds good, compression should barely change it. A useful test is to set the compressor so it only touches the loudest words, reduce the amount until the change is hard to hear, and then compare the two versions. If you cannot tell them apart on headphones and a phone, the compressor is doing its job quietly. If the voice has become smaller or tighter, the setting has gone too far.
Keep a copy of the uncompressed version. Compression is easy to add later, but hard to remove once the voice has been squeezed and the original detail has been lost.
Attack and release are easier to understand by listening than by reading about them. Set a fast attack on a spoken phrase and listen for the start of each word being softened. Then slow the attack and listen again, until the consonants feel present. Repeat with the release, and notice how a quick release makes the voice seem to breathe between words, while a slow one smooths the whole phrase.
Compressors differ in their character, and the same settings will sound different on different units. Use the principles above as a starting point, and adjust by ear. Compression is a matter of taste and of the material, and the best setting is the one that makes the voice clear and natural.
The clearest way to learn compression is to listen to a voice with and without it, at the same level. The LUFS meter can show the loudness of the final file, and the Vocal Checker shows clipping, noise and level problems in a recording you own. For the wider level basics, read the dynamic range post in this series.
Spot clipping, noise, rumble and level problems. Runs in your browser, nothing is uploaded.
It reduces the loud parts, which lets the level be raised. It is not a loudness button.
A gentle ratio, around 2 to 1 or 3 to 1, is a sensible start for most voices.
The compression is too strong, or the ratio is too high. Reduce the amount and listen.
Usually it is safer to record without heavy compression, and apply gentle compression afterwards.
