Free · private · in your browser

Compress the dynamic range of video audio

Choose a preset, compare the sound, refine the balance, then export.

How it works

ImportChoose source
EditSet the change
PreviewReview result
ExportCreate file

InputMP4, WebM, MOV and M4V videoOutputMP4 or WebM video

MP4 · WebM · MOV · M4VChoose a videoBalance quiet and loud moments.

When a recording has quiet passages you have to lean in for and loud moments that make you flinch, the problem is dynamic range. A compressor reduces the distance between the two so the whole thing sits comfortably at one volume — which is why nearly every piece of broadcast speech you have ever heard has been through one.

Threshold, ratio, attack and release

The threshold is the level above which compression starts, and the ratio is how firmly it acts: 2:1 is gentle, 4:1 is standard for speech, and 8:1 or higher is heavy-handed. Attack sets how fast it responds — fast attack catches sharp consonants but can dull the front of words, slow attack preserves punch and lets more through. Release sets how quickly it recovers; too fast and you hear the background level pumping up and down between phrases, too slow and it never lets go. For dialogue, a moderate ratio with a medium attack and release is a safe starting point.

Make-up gain is where the loudness appears

Compression only ever turns things down, so on its own it makes a track quieter. Make-up gain raises everything afterwards, and since the peaks have already been reduced, the quiet parts come up further than the loud ones — which is the whole point. That is also where over-compression shows: pushed too far, breaths and room noise rise to the same level as speech, and the result sounds airless and fatiguing. If a recording starts sounding like it is being squeezed, reduce the ratio before touching anything else.

Questions, answered

What compression settings work for speech?

A ratio around 3:1 or 4:1, with the threshold set so the loudest words are reduced by a few decibels, and medium attack and release. Then use make-up gain to bring the level back.

What is the difference between a compressor and a limiter?

A compressor reduces dynamic range gradually across the whole track. A limiter is an extreme compressor acting only at a ceiling, used to stop isolated peaks rather than to shape the sound.

Why does my audio sound like it is breathing?

The release is too fast, so the background level audibly rises between phrases. Lengthen it, and reduce the ratio if it persists.

Should I compress before or after normalising?

Before. Compression changes the loudness, so normalise afterwards to land the finished track on your target level.

Do I need a compressor at all?

Not always. Evenly recorded speech from a close microphone may only need normalising. Compression earns its place when the level varies a lot within one recording.

Compress the dynamic range of video audio

Threshold, ratio, attack and release

Make-up gain is where the loudness appears

Questions, answered

Useful next steps