You bounced your track, dragged it into a reference playlist next to a major-label single, and yours sounds small. So you slam a limiter on the master bus, push the gain until the meter screams, and tell yourself that's how the pros do it. Then you read that Spotify is just going to turn it down anyway, and now you have no idea what number to chase.
The short answer: Aim for roughly -14 LUFS integrated with a true-peak ceiling of -1 dBTP, and stop worrying about being "loud enough." Streaming platforms normalize playback, so an over-compressed master gets turned down to the same level as a dynamic one — it just arrives more damaged.
Why loudness stopped being a competition
For most of the 2000s, louder genuinely won. A hotter master jumped out on radio and on shuffle, so engineers crushed dynamics to claw out every last decibel. That era is over.
Every major streaming service now normalizes playback to a target loudness. When a listener has normalization on, the platform measures your track's integrated loudness and adjusts the playback gain so it sits at the same perceived level as everything else.
The reference targets most people quote are:
- Spotify — around -14 LUFS
- Apple Music — around -16 LUFS
- YouTube — around -14 LUFS
- Amazon Music — around -14 LUFS
- Tidal — around -14 LUFS
The critical detail: these services only turn loud masters down. If you deliver a brick-walled -7 LUFS master, Spotify pulls it down roughly 7 dB to hit its target — and now your squashed, lifeless track plays back at the exact same volume as a master that kept all its punch. You took the damage for free and got nothing in return.
What the numbers actually mean
LUFS stands for Loudness Units Full Scale — a measurement of perceived loudness over time, weighted to reflect how human hearing works. "Integrated" LUFS is the average across the whole song. That's the number normalization cares about.
Two other meters matter at the master stage:
- True peak (dBTP) — the actual highest peak after digital-to-analog reconstruction. Keep this at -1 dBTP. Lossy codecs like AAC and Ogg Vorbis can push peaks higher during re-encoding, and that overshoot causes clipping on playback. The -1 dB of headroom absorbs it.
- Loudness range (LRA) — how much the loudness moves across the track. There's no target to hit; it's a sanity check that you haven't flattened everything into a wall.
The practical recipe at the master bus is straightforward: get your mix sounding finished and balanced first. Then put a true-peak limiter last in the chain, set the ceiling to -1 dBTP, and bring in just enough gain to land somewhere around -14 to -10 LUFS integrated — watching the sound, not the number, the whole way.
The myth that won't die
The number worship is the problem. Producers treat -14 LUFS like a pass/fail line on an exam, convinced that anything quieter is broken and anything louder is "competitive."
The misconception: "I have to hit exactly -14 LUFS or my track is too quiet." Wrong. -14 isn't a quality bar — it's a normalization reference. If your master lands at -16 LUFS, Spotify simply turns it up to match, and it sounds identical in level to a -14 master with more breathing room. Quieter is fine. Louder just gets turned back down, and arrives squashed.
Once you internalize that, the whole exercise changes. You're not hitting a target. You're choosing how much dynamic life to keep, knowing the platform handles the level.
The other camps
Plenty of working engineers land somewhere other than -14, and they're not wrong — they're optimizing for a different listener. Here's where reasonable people split.
The conservative camp (-16 LUFS). Some mastering engineers target the quietest common reference, around -16 LUFS, so the master never gets turned down on any platform and keeps maximum dynamic range everywhere. The trade-off: if a listener has normalization off, your track may feel a touch soft next to louder releases.
The loud camp (-9 to -7 LUFS). For club, hip-hop, trap, and EDM, loudness still does real work. Bass weight and aggressive limiting are part of the genre's identity, and a meaningful chunk of that audience listens with normalization off — on SoundCloud, in DJ sets, through systems where the loudest file wins. These engineers knowingly push to -9 or even -7 LUFS, trading dynamics for impact because their audience rewards it.
The multi-master camp. The pragmatic answer many pros land on: don't pick one. Deliver a -14-ish streaming master and a separate, louder master for clubs, CDs, or platforms without normalization. Same mix, different limiting decisions, each tuned to where it'll actually be heard.
The honest takeaway is that the right loudness is genre- and destination-dependent. A singer-songwriter ballad and a festival banger should not chase the same number, and any engineer telling you otherwise is selling a one-size formula.
How to do it without obsessing over the meter
Forget the meter until the very end. If you mix into a limiter and let the loudness target drive your decisions, you'll over-process to chase a number and wreck the dynamics that make the song feel alive.
A clean workflow:
- Mix to taste with headroom. Leave your mix bus peaking around -6 dBFS before you touch any final limiting. Balance, tone, and arrangement do far more for perceived loudness than a limiter ever will.
- Add the master limiter last. Enable true-peak mode, set the ceiling to -1 dBTP.
- Push gain by ear, not by number. Bring it up until it feels finished and the loudest sections stay punchy. Check where you landed — anywhere from -14 to -10 LUFS is healthy for most modern music.
- A/B against a reference at matched loudness. Use a plugin that level-matches, or manually gain-match, so you're comparing tone, not volume. This is the single most useful habit in mastering.
- Trust your ears over the readout. If -12 LUFS sounds great and -9 sounds crushed, the answer is -12. The meter doesn't get a vote on whether your track feels good.
A reference track is still your best tool — but only when you match its level to yours before comparing. Almost every "my mix sounds smaller" complaint is really just the reference playing louder. Match the loudness and the gap usually disappears, which tells you the problem was never loudness in the first place.
The real skill isn't hitting -14. It's making a master that sounds full, controlled, and emotionally right at whatever level the platform plays it. Chase the sound, set a sane true-peak ceiling, and let the algorithm handle the rest. The records that hold up next to major-label releases aren't winning on LUFS — they're winning on balance, arrangement, and restraint, and those are the things no meter will ever measure for you.