What does normalizing audio do?

This tool measures the file's highest sample, then applies the one gain that puts it on your target. A peak of -12 dB normalized to -1 dB gets +11 dB, and every other sample moves by that same 11 dB.

Will it make a quiet part louder compared with a loud part?

No. Every sample gets the same gain, so the gap between quiet and loud passages stays exactly as it was. For that you need dynamic range compression or a limiter. The Volume Booster's limiter can hold peaks down while raising the rest.

Does this normalize to LUFS?

No. It does peak normalization and measures sample peaks only. It does not measure LUFS or true peak, so it cannot promise you -14 LUFS for Spotify or -16 LKFS for Apple Podcasts. Use a loudness meter in an audio editor for those.

My file is already at the target. What happens?

The page warns you when the gain would be under 0.1 dB. You can still run it, but the level stays the same and the file is encoded again, which for MP3 output means another round of lossy compression. If it still sounds quiet, try the Volume Booster.

No. It is read and processed in your browser and is not uploaded. Limits are 100 MB and 30 minutes. Only the first two channels of a surround file are kept. The page sends ordinary analytics such as the target you chose, never the file or its name.

Audio Normalizer

Set an audio file's highest peak to the level you choose, -1 dB by default. See the measured peak and the gain first. This is peak normalization: the whole file moves by the same amount, so quiet parts stay quiet relative to loud ones.

Drop your recording here or click to browse

MP3, M4A, WAV, AAC, FLAC, OGG, OPUS, WEBM or MP4, up to 100 MB and 30 minutes.

The file is processed in your browser and is not uploaded.

How to normalize an audio file

1
Drop the file
MP3, M4A, WAV, AAC, FLAC, OGG, OPUS, WEBM or MP4, up to 100 MB and 30 minutes. It is read in your browser.
2
Read the peak
The page measures the highest sample in the file and shows it in dB below full scale (dBFS), with the average level beside it.
3
Pick a target
-1 dB is the default; -3, -6 and 0 are there too. The panel shows the gain before you commit, and warns you if the file is already there.
4
Save it
Normalize, compare before and after, and download an MP3 or a 16-bit WAV.

Want it louder than its peak allows? That is the Volume Booster, with a limiter.

Peak normalization versus loudness normalization (LUFS)

This tool does peak normalization. It finds the single highest sample, works out the gain that puts that sample on your target, and applies that same gain to every sample. The sum is target minus measured peak: a file peaking at -12 dB, normalized to -1 dB, gets +11 dB.

Loudness normalization measures how loud the whole program sounds, in LUFS (the ITU-R BS.1770 method, which weights frequencies and ignores near silence), and sets the gain from that. Two files with the same peak can differ a lot in LUFS: a single loud clap sets the peak for an otherwise quiet talk. This page does not measure LUFS and cannot hit a LUFS target.

Because every sample gets the same gain, the balance inside the file does not change. A quiet passage stays exactly as far below the loud one as before. If one speaker is faint and another is close to the mic, the gap survives. Evening that out takes dynamic range compression or a limiter, which this page does not apply. The Audio Compressor on this site is a different thing: it makes the file smaller, not the levels closer.

What normalizing to -1 dB does

Each row is a file with that peak and a quiet passage 30 dB below it. The gap between them stays 30 dB in every row.

Measured peak	Gain applied	Peak after	Quiet passage before	Quiet passage after
-18.0 dB	+17.0 dB	-1.0 dB	-48.0 dB	-31.0 dB
-12.0 dB	+11.0 dB	-1.0 dB	-42.0 dB	-31.0 dB
-6.0 dB	+5.0 dB	-1.0 dB	-36.0 dB	-31.0 dB
-1.0 dB	0.0 dB	-1.0 dB	-31.0 dB	-31.0 dB
0.0 dB	-1.0 dB	-1.0 dB	-30.0 dB	-31.0 dB

A file already above the target gets a negative gain, so normalizing can also turn a file down.

Which target to choose

-1 dB: the default
Leaves a little room for an MP3 or AAC encoder, which can push peaks slightly above the source. Spotify, Apple Podcasts and EBU R 128 all set their true-peak limits at -1 dB.
-3 dB: more room
A sensible choice before more editing, such as adding music, or when several clips will be mixed together.
-6 dB: lots of room
For a file you will process further, or mix under another track. It will sound quieter than the other targets.
0 dB: full scale
The highest sample touches the very top. Reasonable for a 16-bit WAV you will not encode again; risky for MP3, whose decoded peaks can go over and clip.

Choosing between the two tools? The Volume Booster can go past this limit, and the Audio Enhancer reduces steady background noise in a voice recording.

Published loudness targets (not what this tool sets)

These are loudness figures in LUFS or LKFS, two names for the same scale. Setting a peak target does not get you to any of them; they are here so you can tell the two apart.

Who	Loudness	True peak	Applies to
Spotify	-14 LUFS	Below -1 dB TP	Music, adjusted during playback
Apple Podcasts	About -16 LKFS, ±1 dB	Not above -1 dB FS	Podcast episodes, recommended
EBU R 128	-23 LUFS	At most -1 dBTP	European broadcast programs

As published by each on November 4, 2026. This page measures sample peaks, not true peaks, which can sit slightly higher between samples.

Recording the meetings yourself?

The ParrotNotes app records your meetings and interviews, transcribes them and summarizes them.

Get Started Free

Related tools