Platform tips:Why your music sounds quieter after you upload

You mix your video, it sounds great, you upload it, and on the platform it sounds quieter or flatter than it did in your editor. This is usually loudness normalisation.

What normalisation does

Most large platforms measure the loudness of each upload and turn playback down for videos that are louder than their reference level. YouTube's reference is commonly measured at around -14 LUFS. A video mixed much louder than that gets turned down. It does not get any louder than everyone else, it just loses dynamics it did not need to lose.

Why pushing loud backfires

If you squash your mix with heavy limiting to make it loud, normalisation turns it down to the same level as everyone else, but it keeps the squashed, flat sound. A mix with natural dynamics at a moderate level often sounds better after upload.

A practical target

• Mix your voice so it is consistently clear.
• Keep the music under it at a sensible level. Ducking music under voice has a starting point.
• Measure the whole video's loudness on export. Around -14 LUFS is a reasonable target for YouTube.
• Keep true peaks below around -1 dB to avoid distortion after the platform re-encodes your audio.

Start with a good source file

Normalisation and re-encoding are kinder to clean source audio. Import your music file as it is, without converting it, and export your video at a high audio setting. WAV or MP3 for video editing explains why.

Check on the platform

After uploading, listen on the platform itself, on a phone and on headphones. That is what your audience hears.

For podcast levels, see Mixing podcast music to loudness targets.