If you’ve wondered why your track sounds quieter than commercial releases – or why cranking the limiter doesn’t help on Spotify – the answer is LUFS and loudness normalization.
🔧 Free tool: try our LUFS Loudness Targets.
What LUFS means
LUFS (Loudness Units Full Scale) measures perceived loudness over time, much closer to how we actually hear than peak meters. Integrated LUFS is the average loudness of your whole track – the number that matters most for release.
Why streaming changed the game
Spotify, Apple Music and YouTube normalize playback to a target (commonly around -14 LUFS). Master much louder than that and they simply turn it down – so over-compressing just costs you dynamics and punch without making you louder to the listener.
How loud should your master be?
- Streaming-friendly: roughly -14 LUFS integrated is a safe, common target.
- Genre matters: loud genres often master hotter; acoustic/jazz stay more dynamic.
- Prioritise sound over numbers: a punchy, dynamic master beats a crushed loud one.
The three LUFS readings you’ll actually use
Most loudness meters show more than one number, and it helps to know which one to watch. Integrated LUFS is the whole-track average and the figure you compare against a release target. Short-term LUFS measures loudness over a rolling three-second window, which is handy for checking whether a chorus or drop is jumping out as intended. Momentary LUFS reacts almost instantly and is useful for spotting transient peaks. For mastering decisions, focus on integrated; use short-term to sanity-check the loudest sections of the song.
One more reading worth understanding is true peak, measured in dBTP (decibels true peak). Even when your samples sit below 0 dBFS, the analogue waveform reconstructed by a listener’s converter can overshoot and clip – especially after lossy encoding to formats like AAC or Ogg. Leaving a true-peak ceiling of around -1 dBTP gives the encoder room to breathe and keeps the master clean across platforms; if the concept is new to you, our guide to true peak breaks it down in detail.
Why platform targets aren’t all the same
The -14 LUFS figure gets quoted as if every service agreed on it, but each platform sets its own playback target and handles masters differently. Knowing how they behave stops you second-guessing your meter.
- Turning down vs turning up. Most services only attenuate – if your master is louder than their target, they pull it down. Some also apply positive gain to bring quieter masters up toward the target, but they do this carefully so they don’t introduce clipping, which is another reason to respect your true-peak ceiling.
- Optional loudness modes. A few platforms let listeners switch normalization off entirely, or offer a louder “loud” setting. You can’t control which mode someone uses, so mastering for the normalized case keeps you safe for the majority.
- One master, many destinations. Because the louder you go the more a normalizing platform simply claws back, a single sensible master that sounds good around the common target will translate well almost everywhere. There’s rarely a need to render a separate, hotter file per service.
How to hit your LUFS target without crushing the mix
Loudness is the last thing you should chase, not the first. Work through it in order:
- Get the mix balanced first. A well-balanced mix is already most of the way to a competitive master. If the low end is muddy or the vocal is buried, no amount of limiting will fix it.
- Set the tone, then the level. Make your EQ and dynamics moves for the sound you want, then bring up the level with a limiter or maximiser at the very end of the chain.
- Push the limiter gently. Add gain into the limiter a little at a time and listen, rather than slamming it. A few decibels of gain reduction on peaks is normal; consistently burying the needle usually means you’ve gone too far.
- Check on real systems. Reference your master against a commercial track you like in the same genre, ideally with both played at matched loudness so you compare tone and punch rather than volume.
Common LUFS mistakes
A few habits trip people up again and again:
- Mastering to a number instead of the music. -14 LUFS is a guideline, not a law. Some tracks sound right a little quieter and more dynamic; others suit a hotter master. Let the song lead.
- Ignoring true peak. Hitting your loudness target but leaving peaks at 0 dBFS invites distortion once a platform re-encodes the file.
- Forgetting where loudness starts. Clean gain structure earlier in the chain makes mastering far easier. If you’re fighting noise or clipping, the problem is usually upstream in the mix.
- Comparing at different volumes. Louder almost always sounds “better” for a second. Always level-match before you judge whether a change actually improved the master.
Frequently asked questions
Should I master louder than -14 LUFS so my track stands out?
Not for streaming. Platforms that normalize playback will simply turn a louder master down, so you gain nothing in perceived loudness and lose dynamic range in the process. If anything, a slightly more dynamic master can sound bigger and more open once everything is played back at a matched level. Mastering hotter mainly makes sense for contexts that don’t normalize, such as some club or download use.
Does a lower LUFS number mean my master is too quiet?
No. LUFS is negative because it’s measured relative to full scale, so -14 LUFS is louder than -20 LUFS. A lower (more negative) reading simply means a more dynamic, quieter average. What matters is whether the master sounds full and controlled, not whether it sits at one exact figure.
Do I need a paid LUFS meter?
Not necessarily. Many DAWs include a loudness meter, and there are capable free options as well. The key features to look for are integrated, short-term and momentary readings plus a true-peak display. Once you can see those, the meter you use matters far less than how you respond to it.
Where should I measure LUFS – before or after my limiter?
Measure at the very end of your chain, after the limiter or maximiser and any final dither, because that’s the signal the listener actually receives. Place the meter on your master bus as the last item so your integrated reading reflects the finished file. If you measure earlier in the chain you’ll see a quieter number that doesn’t match what gets exported.
Should I render the whole song to get an accurate integrated reading?
Yes. Integrated LUFS is an average across everything that plays, so a reliable figure needs the full track from start to finish, including quiet intros and outros. Letting the meter run over the entire song – or measuring the bounced file – gives you the number a streaming platform will see. Checking only a loud chorus will read hotter than the real integrated value.
Practical takeaway
Use a LUFS meter, aim for a sensible target, and don’t sacrifice dynamics chasing loudness. This is part of the bigger picture in what is mastering, and starts with leaving headroom in your mix.
