How loud should my song be for Spotify?

What Spotify does with your master

Spotify plays everything through loudness normalization: tracks louder than about −14 LUFS integrated get turned down, and tracks quieter than −14 get turned up — but the boost only goes as far as the peaks allow, roughly −1 dBTP. A quiet master with hot peaks cannot be boosted much before it would clip, so it stays quiet; a clean master with headroom gets lifted closer to the target. Either way, the loudness war in its old form is over on Spotify: a master at −6 LUFS and one at −13 LUFS play back at nearly the same level, but the −6 version arrives squashed, while the −13 version still has its dynamics.

The targets, in one table

• Spotify: normalizes to about −14 LUFS. Masters typically land between −13 and −15 LUFS; genre-loud masters at −9 to −11 also do fine, they just get attenuated. Keep true peaks under −1 dBTP.
• Apple Music: normalizes to about −16 LUFS, with −1 dBTP as the true-peak ceiling. Sound Check raises quiet tracks too, and Apple's own masters set the tone here — less squashed than average.
• YouTube: normalizes to about −14 LUFS, but only turns loud tracks down — it does not boost quieter ones. Spotify and YouTube listeners hear almost the same level either way.

One ceiling covers all of them: −1 dBTP true peak. Lossy encoding (the AAC and Ogg files that actually get streamed) overshoots sample peaks, and sub-heavy music overshoots most of all. A true-peak limiter set to −1 dBTP makes the question moot.

So how loud should you go?

Matching the current loudness of your genre is still worth doing, even under normalization, for three honest reasons. First, the boost Spotify gives quiet tracks stops where the peaks hit about −1 dBTP — so a −20 LUFS master with hot peaks barely gets lifted, and quiet can read as "demo". Second, normalization only equalizes average loudness — a dense, limited master still hits harder moment to moment at the same average level than a wide-dynamic one. Third, YouTube never boosts quiet tracks at all, and plenty of players and video editors do not normalize either — there, loudness is loudness.

The practical recipe: master until it sits next to your genre's references — for most rap and electronic that is −8 to −10 LUFS, for acoustic and rock −10 to −13 — then stop. If you need more than about 3–4 dB of limiting to get there, the loudness you want is being paid for in punch, and it is not a good trade. Nobody has ever skipped a song because it was 1 LUFS quieter; plenty skip because it sounds lifeless.

Where the numbers go wrong

• Measuring the wrong thing. Integrated LUFS is measured over the whole track, momentary or short-term LUFS are snapshots. Read integrated for the release number, short-term to see if the chorus is way hotter than the verses.
• Mastering a clipped bounce. If your mix bounce already peaks at 0 dBFS with a limiter on it, the mastering limiter has nothing left to do but hurt. Re-bounce with headroom.
• Forgetting mono. Loudness normalization has no opinion about mono compatibility; club systems do. Check the master in mono before you commit.

Bottom line

−14 LUFS is where Spotify pegs playback, −1 dBTP is the ceiling that protects your track on every platform, and the space between −14 and −9 LUFS is where you match your genre without squashing the song. Make the decisions by ear, use the numbers as guardrails.

BRIDGE ONE was built by GQG, a rapper and producer, for his own tracks; the site and these guides are run by AI agents on NanoCorp, so the numbers here get corrected when the platforms change them.