Short answer: for most streaming, aim for around -14 LUFS integrated with your true peak no higher than -1 dBTP. Streaming services play every track at a set loudness and turn louder masters down to reach it, so pushing for maximum loudness past that target does not make you louder than anyone else. It just trades away your dynamics.
The numbers here are approximate and each platform revises them over time, so treat them as a target zone, not a fixed rule. They are written as integrated LUFS, the loudness of the whole track averaged from start to finish.
LUFS is the standard unit for how loud a track actually sounds to a listener. It stands for Loudness Units relative to Full Scale, and unlike a peak reading it measures loudness the way ears hear it, averaged over time rather than at a single instant. It is the number streaming platforms read off your track to decide how loud to play it.
The figure that matters for mastering is integrated LUFS, a single value for the whole track from beginning to end. A more negative number is quieter and a number closer to zero is louder, so a master at -8 LUFS is a lot louder than one at -16 LUFS. When people ask how loud a master should be, this integrated number is what they mean.
LUFS is not the same as peak level. Peak, and true peak, measure the highest instantaneous point the signal reaches, which is about avoiding clipping. LUFS measures the felt loudness across the whole track. You care about both, but they answer different questions, and the two sections below take each in turn.
Aim for around -14 LUFS integrated with a true peak at or below -1 dBTP. That is close to where most services set playback, so a single master travels well across all of them without a separate version for each. Apple Music sits a little quieter at around -16 LUFS, but it is near enough that you do not need to master twice. Here is roughly where the main platforms land.
| Platform | Approx. target | How it adjusts |
|---|---|---|
| Spotify | around -14 LUFS | Turns loud masters down, lifts quiet ones up within limits. Listeners can pick -19, -14 or -11. |
| Apple Music | around -16 LUFS | Turns loud masters down and lifts quiet ones up. |
| YouTube | around -14 LUFS | Turns loud masters down only, does not lift quiet tracks up. |
| TIDAL | around -14 LUFS | Turns loud masters down only. Normalizes by album. |
| Amazon Music | around -14 LUFS | Turns loud masters down only. |
| SoundCloud | around -14 LUFS | Turns loud masters down only. |
The pattern is that music streaming clusters around -14 LUFS, with Apple Music the quiet outlier near -16. Because these are moving targets and each service tweaks its own setup, treat them as a zone rather than exact figures. A master at around -14 LUFS with a -1 dBTP ceiling is a safe default that works everywhere, and you only reach for something louder when a genre asks for it.
No, not for streaming. The platform turns a loud master back down to its target, so pushing past around -14 LUFS does not make you louder than the track before you. It just means you gave up dynamic range for nothing. Loudness normalization exists precisely so that no track wins by being hotter.
Here is what happens in practice. Say you crush a master to -8 LUFS to make it feel loud. Spotify reads that, sees it is about 6 dB over its target, and turns the playback down roughly 6 dB so it lands at the same level as a track mastered at -14. Now it plays back at the same loudness as everyone else, but with less punch and less space left in it, because you squashed the life out of it to get that number up. The quieter, more dynamic master turned up to the same level sounds more open.
Streaming platforms turn loud masters down. Your loud master does not play louder than anyone else, it plays at the platform's target like everything else, just with the dynamics already squeezed out of it. Chasing loudness past the target buys you nothing and costs you the movement in your track. And some services, including YouTube, TIDAL and Amazon Music, only turn loud tracks down and never lift quiet ones up, so a master that is too quiet simply stays quiet there.
Loud is a genre choice, not a loudness win. If your style expects a dense, hard-hitting sound, some club, hip hop or metal for instance, master for that sound and accept that it will be turned down on playback. That is a deliberate artistic call. What does not work is reaching for loudness on its own, hoping to stand out, because the normalization step erases the advantage before the listener hears it.
Keep true peak at or below -1 dBTP so the file does not clip when it is turned into a streaming codec, and leave your dynamics room to breathe instead of crushing everything up against the ceiling. These are the two things loudness normalization cannot fix for you, so they are worth getting right at the master.
True peak measures the real peak the signal reaches between samples, the inter-sample peaks, which can sit higher than the sample peaks a normal meter shows. When a track is encoded to a lossy format like AAC or Ogg Vorbis for streaming, those hidden peaks can push into clipping. Holding true peak to -1 dBTP leaves about 1 dB of headroom to absorb that, which is why it is the common default. If you deliberately master louder than -14 LUFS, lower the ceiling a little further, to around -2 dBTP, since louder material is more prone to encoding distortion.
Dynamics are the gap between the loud and quiet moments of your track, the movement in it. Loudness is the average level; dynamics are how much it rises and falls around that average. Normalization quietly rewards keeping them, because a track with dynamics turned up to the target sounds open and alive, while a squashed one turned down to the same target sounds flat. So the master that travels best is not the loudest one, it is the one that hits the target with its dynamics intact.
Aim for around -14 LUFS integrated with a -1 dBTP true peak ceiling, and only go louder when a genre genuinely calls for it. Measure with a loudness meter on your master bus, set the loudness first, then check the true peak and pull the ceiling down if it is over. These are approximate figures that platforms revise, so aim for the zone rather than a decimal place.
All of this is the last step. It only matters once you have music worth mastering, and that is the harder part. The targets on this page are simple to hit; having something to point them at is the work.
The Collection is our three instruments together, ARGISH, SILT and REHEAT, with three separate licence keys. They are for the first part of this page, making the music: ARGISH is a self-playing drone synth, SILT is a tape-loop instrument, and REHEAT writes acid lines. You write and record with them, then master the result to the targets above, around -14 LUFS with a -1 dBTP ceiling, and whatever you make with them is yours to release. Every one ships as an Audio Unit, a VST3, an AAX for Pro Tools and a standalone app on macOS, signed and notarized, and as a VST3 on Windows.
You can hear one free in your browser first, no install and no account, then decide.
We write these when there is something worth writing down. One email when a new one lands or a new Tunary instrument ships. No newsletter, no schedule.
Tunary