Audio too quiet? How to make it louder without wrecking it
Updated · 4 min read
A lecture you can barely hear, a voice memo recorded from across the room, a clip that's much quieter than everything else on your playlist. There are two ways to turn audio up — normalizing and changing the volume by a set amount — and they suit different situations. This guide explains the difference, what clipping is, and what to do when the sound is only in one ear.
First, a quick word on decibels and clipping
Digital audio has a ceiling called full scale, written 0 dBFS. Levels are measured in decibels below it: −1 dB is just under the ceiling, −12 dB is well below it. A change of +6 dB doubles the level of the signal; −6 dB halves it.
Anything pushed above the ceiling can't be stored, so its tops are cut flat. That's clipping, and it sounds like crackle or harsh distortion on loud words and drum hits. Turning audio up is easy; the skill is turning it up without clipping.
Normalizing: as loud as possible, safely
Normalize Audio works in two passes. It first scans the whole file to find the single loudest moment (the peak), then turns everything up — or down — by exactly the amount that puts that peak at the level you choose. The Peak level slider goes from −12 dB to 0 dB, and −1 dB is the default: loud, with a little room to spare for players and later conversions.
Because the loudest moment lands on your target, nothing clips. And because each file is measured separately, normalizing several files brings each one's peak to the same level.
Best first step for a quiet recording.
When normalizing doesn't make it louder
Normalizing looks at peaks, not at how loud the recording feels. If a quiet speech recording has one sharp sound — a cough, a door, a microphone bump — that spike may already be near the ceiling, so there's no room to turn the rest up. Two fixes: cut the spike out with Trim Audio if it's at the start or end, or accept a little clipping on that one moment and boost by a set amount instead (next section).
For the same reason, two normalized files can still sound different in loudness: one may be steady speech, the other mostly quiet with a single loud peak.
Changing the volume: a fixed boost or cut
Change Volume applies the same change to every file you add, from −30 dB to +30 dB (+6 dB by default). It doesn't measure anything first, so it's predictable: +6 dB on three files raises all three by the same amount and keeps the balance between them.
Anything that would go past full scale is held at the ceiling. A modest boost on a quiet file is fine; a big boost on a file that's already loud flattens its peaks and can sound distorted. If you hear crackle, use a smaller value — or switch to normalizing.
Which one should you use?
| Situation | Use |
|---|---|
| One recording is too quiet | Normalize, peak level −1 dB |
| Several clips must stay balanced against each other | Change Volume, same dB for all |
| A file is too loud | Change Volume with a negative value, e.g. −6 dB |
| Normalizing barely changed anything | Remove the loud spike, or a small fixed boost with Change Volume |
| You need a precise level for an upload or a mix | Normalize to that peak level |
Both tools save as MP3, M4A or WAV. If you'll edit the result further, save as WAV so you don't add another round of lossy compression (see MP3, M4A and WAV compared).
Sound only in one ear? Mix it to mono
If a recording plays only in your left or right earbud, the microphone was recorded onto one channel of a stereo file. Stereo to Mono mixes both channels into a single one, which plays equally in both ears — and suits voice recordings and podcasts in general. A mono WAV is also half the size of a stereo one.
Mixing down can change the overall level, so do it first and normalize afterwards.
What these tools can't fix
- Uneven levels within one file — a loud interviewer and a quiet guest. Both tools apply one change to the whole file, so the gap between them stays the same.
- Background noise. Turning a recording up turns the hiss and hum up with it.
- Audio that was already clipped when it was recorded. Flattened peaks can't be restored; lower the recording level next time.
All three tools run in your browser, so recordings stay on your device. They use the browser's built-in media support (WebCodecs): recent Chrome, Edge and Safari 17+ work best, and Firefox handles most files.
Questions
What's the difference between normalizing and turning up the volume?
Normalizing measures the file and sets its loudest point to a level you choose, so it never clips. Changing the volume adds the same number of decibels to every file without measuring, which keeps several files balanced but can clip if you boost too far.
What peak level should I normalize to?
−1 dB suits most uses: loud, with a little headroom. Choose a lower level, such as −3 dB, if the file will be mixed with other sounds later.
Why does my boosted audio crackle?
The boost pushed peaks past the digital ceiling, so they were cut flat. Use a smaller boost in Change Volume, or use Normalize Audio, which stops exactly at the peak level you set.