Skip to content

Podcast Loudness Normalizer

Measure your episode's integrated loudness (LUFS), true peak and background noise, then bring it to the right level: -16 LUFS for podcasts, -14 for YouTube or -23 for broadcast. It uses ffmpeg's two-pass loudnorm, with optional gentle noise reduction, and shows the numbers before and after with an A/B player. Your audio never leaves your device.

Files up to 2 hours; add several files to process them one after another.

How to normalise a podcast to -16 LUFS

Quick answer: To normalise a podcast to -16 LUFS, drop the episode, pick the target, press Analyse and normalise and compare and download.

  1. Drop the episode. MP3, WAV, M4A, FLAC, or a video file.
  2. Pick the target. Podcast (-16 LUFS), YouTube (-14) or broadcast (-23). Tick noise reduction only if there's audible hiss or hum.
  3. Press Analyse and normalise. Pass 1 measures, pass 2 applies the gain. The first time, the audio engine downloads (about 32 MB).
  4. Compare and download. Check the numbers, switch between A (before) and B (after) while it plays, then download.

What LUFS means

LUFS (or LKFS) measures loudness the way people hear it, over the whole episode (ITU-R BS.1770). Platforms turn loud files down and some turn quiet ones up, so mastering to their level means your episode sounds the way you mixed it.

Which target?

Apple Podcasts asks for -16 LKFS ±1 dB with peaks under -1 dBFS. EBU R128 sets -23 LUFS and -1 dBTP for European TV and radio. YouTube is widely measured to play at about -14 LUFS, although YouTube doesn't publish a number.

Why two passes?

The first pass measures the whole file. The second applies one steady gain change based on that measurement, so quiet and loud parts keep their balance. Only if that gain would push peaks over the limit does loudnorm switch to gentle dynamic control, and the result tells you when it did.

Frequently asked questions

Is my audio uploaded?

No. ffmpeg runs in your browser. Only the audio engine itself is downloaded from this site, once.

What is true peak?

The highest level the waveform reaches between samples, once played back. Keeping it at -1 dBTP or lower stops distortion when the file is converted to MP3 or AAC.

Will noise reduction ruin my voice?

It's gentle (about 10 dB on steady background noise) and off by default. Use the A/B player to check, and leave it off for clean studio recordings.

How long can the file be?

Free for files up to 10 minutes, one at a time. Creator Pro handles files up to 2 hours and several files in a row.

What does the background noise number mean?

It's the level of the quietest moments between words, in dBFS: under -60 is quiet, -60 to -45 is noticeable, above -45 is noisy. It's a rough guide, not a standard measurement.

Discussion

No questions here yet. Stuck on something, or found a trick worth sharing?