How to Master a Suno Track for Spotify

You finished a track in Suno. In your headphones it sounded huge. Then you uploaded it, opened Spotify, pressed play right after your favorite artist… and your song walked in quiet, thin and somehow small — like it showed up to a stadium gig with a practice amp.

It's not your ears, and it's not bad luck. It's two things stacked on top of each other: Spotify's loudness normalization and the way AI generators build sound. I master Suno tracks for clients, so let me walk you through the whole picture — including the settings from my own sessions, not theory from a manual.

Spotify plays everything at a roughly equal perceived loudness. The default target is -14 LUFS integrated: louder masters get turned down, quieter ones get turned up — but only as far as their peaks allow. There are also two listener settings you don't control: Loud (-11 LUFS, where Spotify applies its own limiter to quiet tracks — you do not want a robot making that decision for your mix) and Quiet (-19 LUFS).

Spotify also asks for true peak below -1 dBTP (below -2 dBTP if your master is louder than -14 LUFS), because converting to lossy formats creates inter-sample peaks — push your file against 0 dB and the encoder adds crackle you never heard in your DAW.

Here's the rule that explains 90% of disappointing uploads: Spotify can change your volume, but it cannot un-bake your damage. Loudness is removable. Clipping, harsh limiting and distortion are printed into the file forever. A crushed track gets turned down to the same level as everyone else — and keeps all of its crunch while losing its punch.

Credit where due: Suno's internal balance is often better than many bedroom mixes. But the file that comes out of the export button has a very recognizable set of ailments:

And listeners hear it. Not consciously, maybe — but blind tests keep showing that even casual audiences clock a raw AI track within seconds and engage with it less. The gap between 'generated' and 'released' is exactly what this article is about.

Before any mastering, squeeze everything you can out of the source. These cost nothing and stack up fast:

Here's where the internet fights. The official Spotify docs say -14 LUFS. Meanwhile, pull any current chart hit into a meter and you'll find -7 to -9 LUFS. Who's right?

Both, annoyingly. Spotify will turn a -8 master down to -14 playback anyway — but density survives the turn-down. A well-made loud master played at -14 still feels more solid than a timid master played at -14. The catch is the phrase 'well-made': if you reach that loudness by slamming a limiter into a harsh, uncleaned AI export, you get a loud and ugly track that Spotify politely turns down to be quiet and ugly.

My personal practice, take it or leave it: I push my masters to -7…-8 LUFS integrated — but only after the harshness is cleaned up, and always with a true-peak ceiling engaged. Here's an actual session, meters don't lie:

And the limiter stage that makes it safe — the Maximizer in IRC 4 Modern mode, output ceiling at -0.30 dB with True Peak detection on. The loudness gain comes last, after EQ cleanup, never instead of it:

Honesty time. Mastering is processing on the finished stereo file — and some Suno problems live deeper than that. If the kick has no transient to enhance, if the vocal is metallic at its core, if the hi-hats are pure fizz — no amount of mastering will fix generation. That's finishing territory: stem separation, artifact cleanup per stem, replacing weak parts with real instruments, then mixing and mastering the rebuilt track. It's the difference between polishing a car and actually repairing the engine.

If you want to go deeper into the generation side first, I've written a full Suno guide and a Suno Studio guide — better source material makes every step above easier.