A voice memo that is barely audible, an interview where one file is much louder than the next, a playlist that makes you reach for the volume between tracks: all of these are loudness problems. Normalizing measures how loud a recording sounds and adjusts it to a target level.
Which target should I choose?
The tool uses the EBU R128 loudness method, through FFmpeg's loudnorm filter. You pick the integrated loudness target:
- -16 LUFS, the default, is the level most podcast platforms recommend for spoken content.
- -14 LUFS is close to what music streaming services and YouTube play back at.
- -23 LUFS is the European broadcast standard for radio and TV.
In every case, the true peak is held at -1.5 dBTP or lower, which leaves headroom so the audio does not clip when it is later converted to MP3 or AAC, and the loudness range is kept around 11 LU, preserving natural variation between loud and quiet passages.
LUFS measures loudness the way people hear it, not just the height of the peaks. That is why two files at -16 LUFS sound equally loud even if one is a quiet talk and the other a punchy song.
Does normalizing change the sound?
For a recording that only needs a volume change, it simply sounds louder or softer. Where the loudest peaks would pass the ceiling, the filter gently compresses them, reducing the dynamic range a little. On most speech that is an improvement; on carefully mastered music it can be audible.
Normalizing cannot remove background noise. Raising a quiet recording raises its hiss along with the voice.
Which formats work?
MP3, WAV, FLAC, AAC, M4A, OGG, WMA and Opus. The audio is saved again in the same format as the input; WAV and FLAC stay lossless.
What happens to my upload?
Normalization runs on our server. The file is uploaded over an encrypted connection, processed in Germany and deleted automatically within two hours.
Is it free?
Yes. Files up to 50 MB are free without an account, within a daily allowance of 20 server jobs; Pro takes up to 500 MB. Normalizing pays off most across a set, such as every episode of a podcast, so they match each other. Drop up to 5 files at once for free, or run a whole folder through the pipeline builder with Pro or credits.
How to normalize audio volume
- Drop your audio file onto the drop zone.
- Choose the target: -16, -14 or -23 LUFS.
- Click Normalize.
- Download the normalized file.
FAQ
Which target for YouTube or Spotify?
-14 LUFS. Both turn louder uploads down to about that level anyway.
Why does my audio sound compressed after normalizing?
The original had peaks that would have exceeded the -1.5 dBTP ceiling, so they were limited. This happens most with very dynamic recordings and the louder targets.
Can it fix distortion?
No. Clipping in the original recording cannot be undone.
Does it work on video?
Extract the audio first with Extract Audio, then normalize it.