How loud should you master for streaming?
There is no single right number, because every platform that normalises turns a loud master down to its own target, so mastering louder than that buys no loudness on playback. Spotify and SoundCloud publish −14 LUFS with a true peak of −1 dBTP. Apple Music publishes a headroom rule rather than a loudness figure, and the other figures you see quoted are reported, not published.
Your report opens with a 3-day trial. Stored on our own servers in the EU, never used to train anything. Without an account it is deleted after 7 days. See a sample report
What the platforms publish
Each figure below says where it comes from. A target marked as reported is the number the field repeats, not one the platform states; the loudness targets page gives each platform’s note and its source.
| Platform | Target | True-peak ceiling | Where the figure comes from |
|---|---|---|---|
| Spotify | −14 LUFS | −1 dBTP; −2 dBTP above the target | Published by the platform |
| Apple Music | −16 LUFS | −1 dBTP | Partly published by the platform |
| YouTube | −14 LUFS | −1 dBTP | Reported, not published |
| SoundCloud | −14 LUFS | −1 dBTP; −2 dBTP above the target | Published by the platform |
| TIDAL | −14 LUFS | −1 dBTP | Reported, not published |
| Amazon Music | −14 LUFS | −2 dBTP | Reported, not published |
| Deezer | −15 LUFS | −1 dBTP | Reported, not published |
| Club / DJ | −8 LUFS | −0.3 dBTP | A working convention |
What happens to a master louder than the target
It is turned down by the difference. A master made at −8 LUFS plays about 6 dB quieter on Spotify than it left the studio, next to tracks that reached the same level without being pushed as hard. The level is given back; the density it took to get there stays.
That is a trade, not a mistake: many genres want that density, and a loud master is a choice you can make on purpose. What it no longer buys on a platform that normalises is a louder playback.
What happens to a master quieter than the target
Here the platforms differ, and most of them do not say:
- Spotify: A quieter master is turned up, but only until its true peak reaches −1 dBTP.
- Apple Music: Whether a quieter master is turned up is not published.
- YouTube: A quieter master is reported not to be turned up.
- SoundCloud: Whether a quieter master is turned up is not published.
- TIDAL: Whether a quieter master is turned up is not published.
- Amazon Music: Whether a quieter master is turned up is not published.
- Deezer: Whether a quieter master is turned up is not published.
So a quiet master is not simply raised to the target everywhere. On Spotify it is lifted only as far as its peaks allow, which is one more reason to leave the headroom the platforms ask for.
True peak, and the stricter ceiling above the target
True peak is the highest level your track reaches when it is played back, between the samples too. Lossy streaming encoders change the waveform slightly, so a master whose peaks sit at the very top can go over after you upload, not before.
Spotify and SoundCloud ask for −1 dBTP, and −2 dBTP for a master louder than their target, because a louder master is more likely to distort when it is encoded. Apple asks for at least 1 dB of headroom.
Genres, and the club
A platform’s target does not change with the genre: a ballad and a drum-and-bass track are turned to the same level. What genre changes is what a finished master usually looks like, how much its level moves, how hard its limiter works, how much low end it carries, and how tightly its bass holds together in mono. RefBooth’s genre profiles judge those, never the platform’s loudness target.
A club is different. Nothing levels the file, so dance masters cut for club systems are commonly made at about −8 LUFS with a true peak of −0.3 dBTP. That is a working convention, not a specification: no platform or standard publishes it.
Check your master
Drop the file you are about to deliver: the report gives its loudness against the platform you pick, its true peak against the ceiling, and what to fix first. In DJ and Studio, each with a 3-day trial. To read more about the unit itself, see what LUFS measures; to read one figure on its own, the LUFS checker and the true peak checker.
Sources
- Spotify: support.spotify.com, checked on 1 November 2026.
- Apple Music: apple.com, checked on 1 November 2026.
- YouTube: support.google.com, checked on 1 November 2026.
- SoundCloud: help.soundcloud.com, checked on 1 November 2026.
- TIDAL: tidal.com, not dated: it states no figure to re-read.
- Amazon Music: developer.amazon.com, checked on 1 November 2026.
- Deezer: support.deezer.com, checked on 1 November 2026.