Stop riding the volume knob
You’ve got three recordings. One was captured on a good mic, one off a laptop, one off a phone in a noisy room. Stitch them together and listeners spend the whole time turning the volume up, then yanking it down when the loud clip hits. Normalizing fixes that. It brings every file to the same perceived loudness, so a podcast, a playlist, or a set of voiceovers all sit at one comfortable level.
Drop in an audio file and the tool measures how loud it actually sounds, then adjusts it to hit -16 LUFS. That’s the EBU R128 target a lot of podcast platforms and broadcasters aim for. It also caps the true peak at -1.5 dB, which leaves headroom so nothing clips or crackles on cheap earbuds.
LUFS, not just “make it louder”
Here’s the thing most simple volume tools get wrong. Peak normalization just finds the single loudest sample and scales everything up until that sample hits the ceiling. A track with one sharp drum hit barely moves, even if the rest of it is whisper-quiet.
LUFS measures loudness the way your ears do, averaged across the whole file. So a quiet, even recording and a punchy, dynamic one both end up sounding equally loud to a listener. That’s why streaming services and podcast hosts use it. Match the LUFS and your content stops being the one episode people complain they can’t hear.
How it runs
This one isn’t browser-side. Your file uploads to our server, FFmpeg runs a two-pass loudnorm analysis (first it measures, then it corrects), and the normalized file comes back for you to preview and download. Two passes take a little longer than a quick volume bump, but the result is far more accurate.
The original file gets wiped within an hour. You download the new version, named normalized_yourfile, and that’s it.
When it’s the right tool
- Podcast episodes where guests recorded on different gear and the levels jump around
- Voiceovers that came out too quiet to sit over background music
- A playlist of clips ripped from different sources that all play at different volumes
- Audiobook or narration chapters that need to match each other
- Anything you’re about to upload to a platform that expects around -16 LUFS
Already compressed your file and just need it leveled? Run it here after. Need a different format too? The Audio Format Converter handles that, and you can normalize either before or after converting.
Common questions
What does -16 LUFS actually mean? It’s a loudness target. LUFS stands for Loudness Units relative to Full Scale, and -16 is the level most podcast platforms recommend for stereo audio. The lower the number’s magnitude, the louder. -16 is a sweet spot: clear without being fatiguing.
Will it ruin the quality? No. Normalizing only adjusts gain, it doesn’t re-record or heavily process the audio. The -1.5 dB peak ceiling actually protects quality by preventing clipping.
Can it rescue a really quiet recording? Mostly, yes. It’ll bring the level up to standard. One catch: if a recording is quiet because the mic gain was low, raising the level also raises the background hiss. Normalization makes it loud, not clean.
What formats can I upload? MP3, WAV, M4A, AAC, OGG, FLAC, and most common audio types, up to 100 MB.
Does it normalize each file to match the others? Each file is normalized to the same fixed -16 LUFS target, so any files you run through separately will line up with each other automatically.
Is there a loudness setting I can change? Not in this version. It uses the broadcast-standard -16 LUFS, which works for the large majority of spoken-word and mixed audio.