Make an audio file louder without clipping
This page lets you increase audio volume on a file that is too quiet: a voicemail you can barely hear, a lecture recorded from the back of the room, or one song that sits lower than the rest of a party playlist. Open the file, raise the gain or click “Normalize to −1 dBFS”, compare the original with the result, then download. Everything happens in your browser: the file is not uploaded.
The tool measures the sound first. The “Readings” table shows the “Peak” (the loudest sample), the “Average level (RMS)” and the number of “Clipped samples”, before and after your settings. Values are in dBFS: 0 dBFS is the most a digital file can hold, and everything else is negative.
How to boost an MP3
- Click “Choose an audio file” or drop the file. MP3, WAV, OGG, M4A/AAC, FLAC or Opus: the page opens anything your browser can play.
- Set the “Gain (dB)” with the slider or type a value from −12 to +24 dB. The +3 dB and +6 dB buttons add a step, 0 dB goes back to the original level.
- Or click “Normalize to −1 dBFS”: the page works out the gain that brings the peak exactly to −1 dBFS.
- Keep the “Anti-clipping limiter (ceiling −1 dBFS)” ticked if you go louder than normalization.
- Compare with “Play the original” and “Play the result” (space bar to start or stop), then click “Download”.
On the waveform, the original is gray and the result is in color; the dotted lines mark −1 dBFS and red shows anything that would go past 0 dBFS.
Gain, normalization and limiter by the numbers
Gain multiplies the amplitude by 10 to the power of (dB ÷ 20). So +6 dB multiplies it by 1.995, almost double; +3 dB by 1.413; +20 dB by 10. Your ear hears roughly twice the loudness for every +10 dB.
| Gain | Amplitude | A −12 dBFS peak becomes |
|---|---|---|
| +3 dB | × 1.413 | −9 dBFS |
| +6 dB | × 1.995 | −6 dBFS |
| +11 dB | × 3.548 | −1 dBFS |
Normalization. A lecture recording peaks at −14.0 dBFS with an average level of −32.0 dBFS. Normalizing applies +13.0 dB (amplitude × 4.467): the peak lands at −1.0 dBFS and the average at −19.0 dBFS. Nothing clips and the waveform keeps its exact shape.
Limiter. A podcast interview has the voice around −28 dBFS but one burst of laughter at −3 dBFS. Normalizing would only add +2 dB. With +10 dB and the limiter on, the voice moves up to −18 dBFS and the laugh, which would reach +7 dBFS, is held at −1 dBFS. The limiter turns the sound down within 2 milliseconds, just ahead of the peak, then lets go over 100 milliseconds. Without it, the table counts samples above 0 dBFS: they would be clipped on export and you would hear crackling.
Before boosting, rule out your hardware with a left right speaker test. To keep only the useful part, start with the MP3 cutter. And to check how loud the room really is, use the decibel meter online.
FAQ
Why does normalizing sometimes make the file quieter?
If the peak is already at −0.3 dBFS, bringing it to −1 dBFS takes −0.7 dB. The file is loud on its peaks but quiet on average: use the gain with the limiter to lift the average level instead.
Is the downloaded file an MP3?
No: it is “WAV 16-bit (lossless)” at the original sample rate, or compressed Opus if your browser offers it. Opus is encoded in real time, so one minute of audio takes one minute.
How far can I push it before it sounds bad?
As long as the limiter only catches a few peaks, you will not hear it. Past +12 dB on dense material the limiter works all the time and the sound flattens out; the note under the table tells you what share of the file is affected.
Does the background hiss get louder too?
Yes. Gain raises the whole signal, noise included: +12 dB lifts hiss at −60 dBFS to an audible −48 dBFS. This tool does not remove noise.