Skip to content

Normalize audio loudness

A loudness normalizer that brings quiet and loud recordings to the same level, measured the way podcast hosts and streaming services measure it. No guessing with a volume slider.

I want Normalize

Settings appear once you add an audio file.

Loudness target

Podcast and voice work uses −16. Spotify, YouTube and Apple Music play at about −14. Broadcast (EBU R128) is −23.

Advanced 2

Ignored by the lossless formats.

Done on your device No upload, no queue, no sign-up.

Bytes sent anywhere: 0 B

Works offline after you’ve used it once.

How to use it

Your audio goes into this browser tab, is measured and brought to the loudness you picked on your device, and comes back even. 0 bytes uploaded

  1. Add your files

    Drop your recordings in. Add a whole set if you want them all at the same loudness.

  2. Check the settings

    Pick the target: −16 LUFS for podcasts and voice, −14 for music on streaming services, −23 for broadcast.

  3. Save the result

    Press normalize. Each file is measured, then levelled with 1.5 dB of headroom so nothing clips.

Why it doesn’t upload

Loudness work means reading every sample of your recording. That happens on your own machine here, so unreleased tracks and private interviews stay where they are.

How to check for yourself

Worth knowing

  • This is loudness normalization, not compression or mastering. It makes a whole file louder or quieter, evenly; it doesn’t even out loud and quiet parts within one file.
  • Already-loud recordings may be turned down, which is the point: −14 LUFS is quieter than most modern masters.
  • The file is re-encoded, so a lossy format loses a little quality. Choose WAV or FLAC to avoid that.
  • Uses the FFmpeg engine, a one-time 31 MB download.

Questions

What loudness should a podcast be?

−16 LUFS for stereo is the usual target for podcasts and spoken word, and what Apple Podcasts and Spotify expect. Mono files are often aimed at −19 LUFS; if in doubt, −16 is safe.

What is LUFS?

Loudness Units relative to Full Scale: a measure of how loud something *sounds* to people over time, rather than its peak level. It is the standard (EBU R128 and ITU-R BS.1770) that streaming services and broadcasters use.

Will normalizing make my audio distort?

No. The result is limited to −1.5 dBTP (true peak), which leaves room for the encoder, so it won’t clip even after being turned into MP3 or AAC.

What’s the difference between this and turning up the volume?

Change Volume shifts the level by the amount you choose, whatever the recording. This measures the recording first and then works out the shift needed to hit the target, so a set of files ends up matching each other.

Read more about this

Next steps

Often used before or after Normalize Loudness. These run on your device too.