How loudness is measured
Loudness is measured in LUFS (loudness units relative to full scale), the ITU-R BS.1770 standard that YouTube, Spotify, TikTok and broadcasters use. Unlike peak level, it weights frequencies the way people hear them and averages over the whole clip, ignoring silence.
Example: a voice clip measured at −28.0 LUFS with peaks at −15.2 dBFS needs +14 dB to reach −14 LUFS. Its peaks then land at −1.2 dBFS, just under a −1 dB ceiling, so no limiting is needed. A clip at −20 LUFS with peaks at −2 dBFS needs +6 dB, which would push peaks to +4 dB; the limiter catches those few peaks so the clip still reaches −14 LUFS without clipping.
Which target to use
| Where it's going | Target |
|---|---|
| YouTube, TikTok, Instagram, Spotify | About −14 LUFS; louder uploads are turned down |
| Podcasts, Apple Music | −16 LUFS |
| TV broadcast (Europe, EBU R 128) | −23 LUFS |
| Twitch clips and stream mixes | Around −14 to −16 LUFS so voice sits above game audio |
Streaming platforms turn loud tracks down to their reference level, so mastering much louder than −14 only loses dynamics. Quiet uploads are often not turned up, which is why normalizing a quiet clip makes the biggest difference. The video picture is copied unchanged; only the audio is processed.
Frequently asked questions
What LUFS should I use for YouTube?
About −14 LUFS integrated, with peaks under −1 dB. YouTube turns louder videos down to roughly that level.
What LUFS is best for TikTok and Instagram Reels?
Around −14 LUFS works well. Voice-led clips that are much quieter sound weak next to other videos in the feed.
Does this change the video quality?
No. The picture is copied without re-encoding; only the soundtrack is adjusted and re-encoded.
What is the difference between LUFS and dB?
dB peak level shows the loudest single moment. LUFS measures how loud the whole clip sounds to people, which is what platforms normalize to.
Was this calculator helpful?