Technique
From the engineer's desk

Why your master sounds quieter on Spotify than your reference

You finished your master. It hits -8 LUFS integrated, peaks at -1 dBTP. You pull up your reference track on Spotify alongside your release, hit play, and your song sounds quieter. You're confused — your master is technically louder than the reference on disk.

Here's what's happening, why it's not a problem with your master, and what to do (or not do) about it.

✦ Michael Venia
✦ March 10, 2026
✦ 6 min read

Spotify normalizes everything to -14 LUFS

Spotify (and Apple Music, and YouTube Music, and the rest of the major streaming platforms) applies loudness normalization on playback. The platform measures the integrated LUFS of every track in its catalog and adjusts the playback gain so all tracks play at the same perceived loudness — Spotify's target is -14 LUFS integrated.

So when your -8 LUFS master plays on Spotify:

  • Spotify reads the track's integrated LUFS as -8
  • Spotify applies -6 dB of gain reduction at playback to bring it to -14 LUFS effective loudness
  • The listener hears your master at -14 LUFS, the same as everything else

When the reference track (mastered at, say, -10 LUFS) plays on Spotify:

  • Spotify reads it as -10
  • Spotify applies -4 dB of gain reduction
  • The listener hears it at -14 LUFS effective loudness

Both tracks end up at the same playback loudness. The louder original master (-8) didn't actually get louder for the listener — it just got reduced more by normalization.

The catch you can't undo

Here's the part that bites artists who push masters too loud: the normalization takes the loudness back, but you don't get the dynamics back.

The process of mastering to -8 LUFS instead of -10 typically involves additional limiting. That limiting flattens transients. Once the transients are flattened in the file, they stay flattened — Spotify's normalization doesn't restore them. You traded dynamics for loudness in the file, and the streaming platform normalized away the loudness, leaving you with flatter dynamics for no playback gain.

A -10 LUFS master (less limiting, more dynamics preserved) plays at the same loudness on Spotify as a -8 LUFS master, but it sounds more open because the dynamics are intact.

This is the trap of the loudness war in the streaming era.

Hear it for yourself

The free Streaming Preview tool lets you hear this with your own song. Drop in your master, add a louder or quieter version of the same song as the reference, pick Spotify, and switch between them — the tool plays both at the level Spotify would. The quieter master sounds bigger and more open. The louder master sounds smaller and more compressed.

This is the comparison that changes minds about loudness war chasing. Once you've heard it level-matched, the louder-is-better intuition disappears.

So why does Spotify still let me upload loud masters?

Spotify accepts any master. They don't reject masters that are too loud — they just normalize them down. You can upload a -6 LUFS master and it plays at the same loudness as a -16 LUFS master on Spotify. The platform doesn't punish you for being loud; it just doesn't reward you for it.

The reason this matters for indie artists: the time and energy spent pushing a master from -10 to -8 LUFS is wasted on Spotify. Worse, it costs you dynamics in the file that other systems (Bandcamp lossless, Apple Digital Masters, direct downloads) do play at the louder level — meaning those listeners hear a more compressed version of your song than they need to.

Master to the loudness the song wants. Let the streaming platform handle the rest.

The quiet-track exception

There's one wrinkle worth knowing about: Spotify only turns a quiet track up as far as its peaks allow.

Spotify raises quiet masters toward -14 LUFS, but it stops when the track's peaks get to about 1 dB below full scale. Spotify's own example: a -20 LUFS track that peaks at -5 dBFS only gets lifted to -16 LUFS, not -14 (Spotify for Artists).

For most indie artists this doesn't matter. The exception is if you're mastering very quietly intentionally (deep ambient, sparse classical, sound design). In that case, know that Spotify may not bring you all the way up to -14 LUFS — your track can sit a bit quieter than other tracks in a playlist. YouTube, TIDAL and Amazon Music don't turn quiet tracks up at all.

What other platforms do

Each major platform has its own normalization behavior:

  • Spotify — -14 LUFS target; turns quiet tracks up only as far as their peaks allow
  • Apple Music — Sound Check on by default for most playback, target -16 LUFS
  • YouTube Music — only turns down tracks louder than about -7 LUFS (independent testing)
  • YouTube proper (video) — about -14 LUFS; only turns tracks down
  • Tidal — -14 LUFS, normalizes whole albums together; only turns tracks down
  • Amazon Music — -14 LUFS target; only turns tracks down
  • Bandcamp — no normalization at all; your master plays at its mastered loudness
  • SoundCloud — normalizes playback; recommends mastering around -14 LUFS

If you're releasing to Bandcamp lossless (where there's no normalization), the louder master does play louder for those listeners. So a strategy of "master to a quieter LUFS for streaming-safe dynamics but accept that Bandcamp listeners hear the master at its file loudness" is internally consistent — the listener experience on Bandcamp matches what's in the file, while the streaming platforms normalize everything to a common floor.

What to actually do about your "quiet" master

Three options, depending on what's bothering you:

  1. Nothing — if your master sounds great when level-matched against the reference, the playback loudness is the streaming platform's problem. Your master is doing its job.
  2. Re-master to the platform target — if you want the file itself to be at -14 LUFS (so Bandcamp / lossless platforms hear it at the same level as streaming), re-master to -14 with no other changes. You'll keep dynamics; you'll lose the file loudness.
  3. Re-master with less limiting — if your -8 LUFS master sounds compressed compared to the reference at -10, pull the loudness back to -10 (less limiting, more dynamics). The Spotify playback will be unchanged; the perceived openness will improve.

Option 1 is right for most artists. Options 2 and 3 are right if you're hearing a specific dynamic problem in the file when you A/B against a reference.

The bottom line

Spotify normalization is not punishing your master. It's preventing the loudness war from making any single track louder than its peers. The master that wins is the one that sounds best after normalization — which is almost always a master with intact dynamics, not the loudest master on disk.

Hear the difference for yourself in the free Streaming Preview tool. Then decide what your master wants.

Ready when you are.

When you're ready, I'll master it for $70 a track and have it back in 24–48 hours.