LUFS Targets for Spotify, Apple Music & YouTube (2026)

Master to -14 LUFS integrated with a -1.0 dBTP true-peak ceiling and your track plays correctly on Spotify, YouTube, Tidal, Deezer, and Amazon Music (keep -2 dBTP there). Apple Music normalizes near -16 LUFS, and TikTok's feed isn't loudness-normalized at all (as of July 2026). One -14 / -1.0 master covers essentially everything.

What LUFS should I master to for each platform?

Here is the full table. "Target" is integrated loudness measured to the ITU-R BS.1770 standard — the whole-track gated average, not a peak meter reading. Where a platform publishes an official number we link it; where it doesn't, the figure is the industry consensus and we say so.

Verified July 2026 — we re-check these quarterly.

Don't make nine masters. Make one at -14 LUFS / -1.0 dBTP (or -2.0 dBTP if Amazon matters to you), and let each platform's normalization do its job. That's the deliverable TrackGleam's free master targets and measures on every render.

What is LUFS, and what does "integrated" mean?

LUFS — Loudness Units relative to Full Scale — is how streaming services measure perceived loudness, defined by the ITU-R BS.1770 standard. Unlike a peak meter, it weights frequencies roughly the way your ear does and averages over time. "Integrated" LUFS is the number that matters for streaming: the gated average across the whole track. Gating means near-silent passages (intros, breakdowns, fade-outs) are excluded from the average, so a long quiet outro doesn't drag your reading down. Short-term LUFS is a rolling 3-second window and momentary is 400 ms — useful while mixing, irrelevant to normalization. When a platform says "-14 LUFS," it means integrated, BS.1770-gated. That's also exactly what TrackGleam reports: ITU-R BS.1770-4 gated integrated loudness plus true peak, measured on every master, on your device.

Why do louder masters get turned down?

Normalization exists so listeners don't ride the volume knob between tracks. The platform measures each track's integrated LUFS and applies a fixed gain offset at playback: a -9 LUFS loudness-war master gets pulled down about 5 dB on Spotify, and all the limiting distortion it paid for that level comes along for free — at the same playback volume as everyone else. Crushing your master past -14 buys you nothing on normalized platforms.

The reverse is not symmetric. Spotify does boost quiet tracks, but only within the track's headroom — its own example: a -20 LUFS track with -5 dBTP peaks only gets lifted to -16, not -14. YouTube, Apple Music, and Tidal don't boost at all (as of July 2026). So an under-loud master genuinely plays quieter on most services — which is why "louder" complaints are usually real, just misdiagnosed. We break down that failure mode in why your song sounds quieter on Spotify.

Which platforms break the -14 rule?

Three exceptions are worth knowing. Apple Music normalizes near -16 LUFS via Sound Check, and only downward — your -14 master simply plays about 2 dB lower there than a -16 one would. That's fine; don't make a separate Apple master. Amazon Music is the true-peak outlier: the consensus ceiling is -2.0 dBTP rather than -1.0, because peaks can rise when audio is re-encoded to lossy formats — the mechanics are in our true peak and dBTP explainer. TikTok's in-feed playback isn't loudness-normalized at all (as of July 2026), so a louder file literally plays louder while people scroll. If a track is TikTok-first, a hotter delivery genuinely helps there — just don't upload that version to the normalized platforms, where it only buys distortion. Deezer's -15 LUFS target is a footnote: one LU inside the -14 cluster, no separate master needed.

How do I hit -14 LUFS without crushing dynamics?

The target is a level, not a texture. You can be at -14 LUFS with punchy, dynamic audio or with a flattened sausage — normalization plays both at the same volume, and the dynamic one sounds better. The metric to watch alongside LUFS is LRA (loudness range): pop and electronic releases commonly sit around 4-8 LU. Getting there is gain staging plus moderate limiting, not maximum limiting.

Being under target is the more common real-world problem. We measured 12 AI-generated music exports from our test library: median -15.2 LUFS integrated (range -16.4 to -12.3), and 8 of 12 sat below the -14 LUFS streaming level — meaning they'd play quieter than commercial releases on most platforms. 3 of 12 also exceeded the -1.0 dBTP ceiling, and the median loudness range was 6.6 LU. So a typical AI export needs lifting to target with true-peak-safe limiting, not crushing. If that's your situation, start with how to master Suno songs free.

How do I measure my track's LUFS for free?

Drop your track into TrackGleam. It masters to -14 LUFS integrated / -1.0 dBTP and shows you the measured numbers — integrated LUFS, true peak, loudness range — computed to ITU-R BS.1770-4 in your browser. Nothing uploads, there's no signup, and the free master is a real downloadable file, not a preview. If you just want the readout, that's free too: measure, compare against the table above, and release when the numbers line up.

Master a track free — no signup, nothing uploads