Normalize Audio
Bring a quiet recording up to a standard level in one step.
Free · runs in your browser · nothing is uploaded

How to normalize audio
- Press Choose audio file, or drop a file onto the page, and wait for the waveform.
- Optionally drag across the waveform to normalize only a part, or type exact start, end or length values under it.
- Under Measure, choose “Peak level” or “Average (RMS) level”.
- Set Target level, from -30 to 0 dB. The default is -1 dB. For Average, type a level such as -18 dB.
- With Average, also check Peak ceiling, from -6 to 0 dB (default -1 dB). It stops the gain from pushing the loudest sample above that point.
- Choose a format in the Save panel and press the Save button. Cancel stops a long job.
- Read the gain in the result panel, play the result in the small player, and press Download.
Zoom with the mouse wheel or the plus button to check a short section first, and use Play to hear the part you selected. With no selection the whole file is measured and saved. The line above the Save button shows the expected length and size. The result saves as MP3, WAV, OGG, FLAC or M4A; MP3 is re-encoded at a constant bitrate.
Normalize a part by selecting it, but note that only the selection is saved. To level two clips to the same target, run each one with the same Measure and Target level and compare the gain shown in the result panel. If the panel says the ceiling held the gain back, raise Peak ceiling, up to 0 dB, or lower the target.
How does audio normalization work?
The tool scans the file to find its level, works out the gain in decibels that moves that level to your target, and multiplies every sample by the same number. A file peaking at -4 dB with a Peak target of -1 dB receives +3 dB. A silent file is returned unchanged.
Because the gain is constant, quiet and loud passages move together and the dynamics stay as they were. Noise rises by the same amount as the sound, so a quiet recording with hiss ends up with louder hiss.
What is the difference between Peak and Average (RMS)?
Peak uses the loudest single sample, so the result touches the target exactly and nothing clips. Average (RMS) uses the overall level, which follows perceived loudness more closely, but a few loud peaks can push past full scale. The Peak ceiling prevents that and holds the gain back when needed, which the result panel reports.
The tool measures sample peaks and average level, not loudness units (LUFS) and not true peak. It does not offer an LUFS target.
Which target level should I choose?
For Peak, the default of -1 dB leaves a little headroom below full scale, which is a safe choice if the file will be converted to a lossy format afterwards, because encoders can add small overshoots near 0 dB. It is also the starting value on the page.
For Average (RMS), the tool has no preset: type the level you want, such as -18 dB, and keep the ceiling at -1 dB. A lower target gives a quieter result, and 0 dB puts the peak at full scale.
What is the difference between normalizing and amplifying?
Amplifying adds a gain you choose, such as +6 dB, whatever the file’s level. Normalizing measures the file first and works out the gain that reaches a target, so two clips of different loudness end up at the same level. For a fixed boost, use the MP3 Volume Booster.
What gain does normalizing apply?
Normalizing measures the file and applies one constant gain. A sine wave with amplitude 0.25 peaks at -12.04 dB, so raising its peak to -1 dB needs +11.04 dB. The table runs the tool on four test tones, including a case where the peak ceiling holds back an RMS target that would have clipped.
| Test signal | Peak before | RMS before | Setting | Gain applied |
|---|---|---|---|---|
| Sine, amplitude 0.25 | -12.04 dB | -15.05 dB | Peak to -1 dB | +11.04 dB |
| Sine, amplitude 0.5 | -6.02 dB | -9.03 dB | Peak to -1 dB | +5.02 dB |
| Sine, amplitude 0.1 | -20.00 dB | -23.01 dB | RMS to -18 dB, ceiling -1 dB | +5.01 dB |
| Sine, amplitude 0.5 | -6.02 dB | -9.03 dB | RMS to -3 dB, ceiling -1 dB | +5.02 dB (held by ceiling) |
Frequently asked questions
Is it good to normalize audio?
It helps when clips that play together have uneven volume, because they end up at similar levels. It cannot remove noise or distortion, and it raises hiss along with the sound.
Can I normalize several files at once?
One file at a time. Open the next file after saving, using the same settings, and each comes out at the same target.
Does it work on mobile?
Yes, in a phone browser. Phones are limited to 150 MB per file and about 15 minutes of decoded 44.1 kHz stereo, against 600 MB and about 60 minutes on a computer.
Is it free, and is the file uploaded?
It is free and needs no account. The file is processed inside your browser and never leaves your device.
What if my audio is already loud?
If the peak already sits near the target, the gain is close to zero and the output sounds almost the same. In Average mode, the ceiling may hold the gain back.
Does it add compression or reduce quality?
No compression is applied: every sample is multiplied by one constant, so the waveform shape is unchanged. WAV and FLAC keep the result exactly, while OGG and M4A are lossy.
Does it work as an MP3 volume normalizer?
Yes. It opens MP3 files and can save the normalized result as MP3 again (re-encoded at 128 to 320 kbps), or as WAV, OGG, FLAC or M4A.
What LUFS should I normalize to?
This tool has no LUFS mode. Use a Peak target of -1 dB or an Average target such as -18 dB, and use a loudness meter if a platform requires LUFS.
Related tools
- MP3 Volume Booster: use it to raise or lower the volume by a fixed number of dB.
- Audio Compressor: use it to reduce the gap between loud and quiet parts.
- Fade Out Audio: use it to end a clip smoothly.
- Silence Remover: use it to remove quiet gaps before normalizing.
- Audio Mixer: use it to set a gain for each track in a layered mix.
- Audio Joiner: use it to combine normalized clips.
- Audio Trimmer: use it to keep only the part you need.
- Free Audio Editor: use it to run Normalize together with other edits.