Normalise audio volume
Even out a recording so quiet and loud parts sit at a sensible level.
Drop your file here
Processed securely and deleted within the hour.
Loudness, not peaks
Most volume tools do peak normalisation: find the loudest single sample and scale everything until it touches maximum. That does almost nothing useful for a recording with one loud cough in an otherwise quiet hour — the cough hits the ceiling, everything else stays quiet, and the file is technically "normalised" and still unlistenable.
This does loudness normalisation to the EBU R128 standard, which measures perceived loudness across the whole recording and adjusts to a target. That is what people actually mean when they say a recording is too quiet, and it is what every streaming platform applies to what you upload.
Choosing a target
−14 LUFS is what Spotify, YouTube and Apple Music normalise to. Delivering at this level means the platform leaves your audio alone.
−16 LUFS is the podcast convention and the default here — slightly quieter, with more headroom for speech dynamics.
−19 LUFS suits spoken word intended for quiet listening. −23 LUFS is the European broadcast standard.
What it cannot fix
Normalising raises the noise floor along with the content. A recording made too far from the microphone becomes a louder recording of a distant voice with audible hiss. It sets the level correctly; it does not improve the capture.
Clipping already baked into a recording is likewise permanent — the peaks were flattened when it was recorded and the waveform above them no longer exists.
Why platforms do this anyway
Uploading something far louder than the target means the platform turns it down, and loudness added by heavy compression is lost while the damage remains. Delivering at the target is how you keep control of how it sounds.
Common questions
How is this different from turning up the volume?
It targets perceived loudness across the whole recording rather than scaling until the loudest sample hits maximum.
What target should I use?
−14 LUFS for streaming platforms, −16 for podcasts, −23 for broadcast.
Will it fix a noisy recording?
No. It raises the noise floor along with the voice. It sets the level; it does not improve the capture.
Can it fix clipping?
No. Clipped peaks were flattened at recording time and no longer exist in the file.
Does it work on video?
Yes, video files are accepted and their audio track is normalised.