Normalize a track to a LUFS target for Spotify, YouTube, or Apple Music, or set a simple peak ceiling. Free, and nothing is uploaded.
The volume normalizer brings a track to a consistent loudness — either to a LUFS target that matches what streaming services expect, or to a simple peak ceiling. The distinction matters more than it sounds: peak normalisation only looks at the single loudest sample, which is why two files can peak identically and still sound nothing alike. LUFS measures perceived loudness the way your ears weigh it, which is what platforms actually turn their volume knob against.
Use it to even out a set of tracks so listeners aren't reaching for the volume between them, to prepare a track for upload without it getting quietly turned down, or to bring a quiet recording up to a usable level. If you want tonal shaping and dynamics control alongside loudness, that's the audio master instead. How to normalize audio volume explains the targets.
LUFS (Loudness Units Full Scale) measures perceived loudness the way listeners actually hear it, not just raw peak level. Streaming platforms normalize every track to a reference LUFS on playback, so a song that's too quiet gets turned up and one that's too loud gets turned down. Normalizing to the right target yourself means your track arrives on the platform at the intended loudness instead of being adjusted in ways you didn't choose.
Match the destination: Spotify and YouTube normalize to about −14 LUFS, Apple Music to around −16, while club and loudness-maximized masters often sit near −9. If you're distributing to streaming, −14 is a safe default. If a track is heading somewhere specific, pick that platform's target so the loudness lands where you expect after their own normalization.
Peak normalization scales the file so its loudest single sample hits a chosen ceiling — useful for maximizing level without clipping, but it ignores how loud the track actually feels. LUFS normalization targets perceived loudness across the whole track, which is what streaming platforms care about. For release-ready loudness use LUFS; for simply raising a quiet file safely to a peak ceiling, use peak.
Turning a track up can push peaks past 0 dBFS and clip. To prevent that, the LUFS mode has a 'prevent clipping' option that caps the true peak at −1 dBTP, reducing the applied gain if hitting the loudness target would otherwise clip. The before/after readout shows both the loudness and the true peak so you can see exactly what happened. True peak is estimated with oversampling to catch inter-sample peaks a simple peak meter misses.
No. Both the loudness measurement and the gain processing happen locally in your browser via the Web Audio API. Nothing is uploaded, and the analysis uses the same ITU-R BS.1770 method the streaming platforms use.