How to normalize audio volume (and fix quiet recordings)
Even out loud and quiet audio to a consistent level, understand peak vs LUFS loudness, and hit the targets streaming platforms expect.
You've cut together three recordings and one of them is half the volume of the others. Or you've finished a podcast episode and it's noticeably quieter than everything else in the listener's feed. Normalizing fixes both — it adjusts the whole file so it lands at a consistent, predictable level.
Normalize an audio file
Load your file
Drop an MP3, WAV or M4A into the Normalize Audio tool.
Let it analyse
The tool measures the existing level and works out how much gain to apply.
Export
Download the normalized file at a consistent loudness.
Peak normalization vs loudness normalization
Peak normalization finds the single loudest moment in the file and raises everything until that peak hits a ceiling — usually just under 0 dB. It guarantees nothing clips, but it says nothing about how loud the track actually feels. One stray handclap can hold the entire recording down.
Loudness normalization measures average perceived loudness over time, in LUFS, and targets that instead. It's what streaming platforms use, and it's the better match for how people actually hear volume. A podcast normalized to a LUFS target will sit comfortably alongside other podcasts; the same file peak-normalized might still feel quiet.
What loudness should I target?
- Podcasts and spoken word: around −16 LUFS for stereo, −19 LUFS for mono.
- Music on Spotify, YouTube and Apple Music: around −14 LUFS — these platforms turn down anything louder anyway.
- Broadcast: −23 LUFS is the standard in Europe, −24 LUFS in the US.
Mastering far louder than the platform target gains you nothing — it gets turned down on playback, and you've sacrificed dynamic range for no benefit.
Normalizing won't fix everything
Normalization raises the whole file uniformly, noise included. If the original recording is quiet because the mic gain was too low, turning it up also turns up the hiss. And if the loud and quiet parts vary wildly within a single recording, normalization won't even them out — it moves the whole thing by one amount. For that you want compression, or simply to re-record with better levels.
The rest of the clean-up pass
- Remove Silence strips long pauses and dead air, which tightens up interviews and voice-overs.
- Volume Audio is the manual alternative when you want to set the level by ear rather than by target.
- Fade Audio smooths abrupt starts and endings.
- Merge Audio joins the finished segments into one file — normalize each part first so the levels match.