Volume Booster

How to use the Volume Booster

1. Choose or drag in the audio file you want to change.
2. Pick how to set the volume: by decibels, by percent, or Normalize peak to make the loudest part reach a set level.
3. Check the numbers under the controls. If they warn about clipping, lower the boost or keep the soft limiter switched on.
4. Click Apply and create WAV and listen to the result in the player.
5. Download the WAV when it sounds right.

How it works

Digital audio is a list of sample values between −1 and +1, where ±1 is "full scale", the loudest a file can store. Changing the volume multiplies every sample by the same factor. In decibels the factor is 10^dB/20, so +6 dB multiplies by about 2.0, −6 dB by about 0.5, and +20 dB by 10. In percent, 200% is a factor of 2. Our ears are not linear, so a sound needs roughly +10 dB before most people call it "twice as loud".

Normalize peak finds the single loudest sample and picks the factor that puts it exactly at your target. A file whose peak is −9 dBFS normalized to −1 dBFS gets +8 dB of gain (×2.51). Because it only looks at the peak, a recording with one loud spike may still sound quiet after normalizing.

If a boost would push samples past full scale they "clip": the tops of the waves are cut flat, which sounds harsh and crackly. With the soft limiter on, the tool leaves everything below −1 dBFS untouched and smoothly bends anything louder so it approaches, but never goes past, −0.1 dBFS, just under full scale (a tanh-shaped curve). It is only applied when the change would actually clip. Heavy limiting still changes the sound, so the panel shows how many samples are affected. The result is saved as a 16-bit WAV file.

Frequently asked questions

How much can I boost without distortion?
The panel shows the current peak. The safe boost is the distance between that peak and 0 dBFS: a file peaking at −8 dBFS can take about +8 dB cleanly. Beyond that the limiter has to round off peaks, and very large boosts will sound squashed.

What is the difference between boosting and normalizing?
Boosting adds the amount you choose. Normalizing works out the amount for you so the loudest point lands at a target level, which is handy for evening out a batch of recordings.

Will this fix a recording that is very quiet and noisy?
It makes everything louder, including background hiss. It can't separate voice from noise, so start from the cleanest recording you have.

Which formats can I open, and how big?
MP3, WAV, M4A/AAC, OGG and FLAC (whatever your browser can play). Decoded audio uses about 10 MB of memory per stereo minute, so files up to about 30 minutes work well on a computer and up to about 10 minutes on a phone.

Is my file uploaded?
No. The audio is processed by your own browser and the new WAV is created on your device.