How Loud Should You Master Your Music?
Contents
-8.4 LUFS
average loudness of Spotify's top 25 tracks — nearly 6 dB above their own -14 LUFS recommendation
The 25 most-streamed tracks on Spotify in 2022 averaged -8.4 LUFS integrated. If the world's biggest records are ignoring the "standard," what should you actually aim for?
What LUFS Actually Measures
LUFS (Loudness Unit Full Scale) is defined by the ITU-R BS.1770 spec. Unlike RMS, which measures raw average power, LUFS applies frequency weighting that models how human ears perceive loudness — more sensitive to midrange (1–5 kHz), less sensitive to extremes. Two tracks at the same LUFS level should sound roughly equally loud to a listener.
Three types of LUFS measurements matter:
- Integrated LUFS — average loudness of the entire track. The number streaming platforms use for normalization.
- Short-term LUFS — 3-second window. Useful for monitoring your loudest sections.
- Momentary LUFS — 400ms window. Real-time snapshot.
And one more critical measurement: True Peak (dBTP) captures the absolute maximum level including inter-sample peaks. Every streaming platform recommends -1.0 dBTP or below to avoid distortion during encoding.
How Streaming Platforms Handle Loudness
Every major platform uses loudness normalization — adjusting playback volume so listeners don't reach for the volume knob between a jazz ballad and an EDM banger.
| Platform | Target LUFS | Notes |
|---|---|---|
| Spotify | -14 LUFS | Users can select Quiet (-23), Normal (-14), or Loud (-11) |
| Apple Music | -16 LUFS | Uses Sound Check; slightly quieter target |
| YouTube | -14 LUFS | Always on; will not turn quiet tracks up |
| Tidal | -14 LUFS | Uses album normalization by default |
| Amazon Music | -14 LUFS | Standard normalization |
| Deezer | -15 LUFS | Slightly quieter than Spotify |
If your master comes in at -8 LUFS and Spotify targets -14, the platform turns your track down by about 6 dB. If you're at -16, Spotify turns it up by 2 dB (though some platforms like YouTube only turn down, never up).
The -14 LUFS Myth
The most common advice is to master to -14 LUFS. It matches Spotify's target and avoids volume reduction. But there are real problems with treating it as universal.
-14 LUFS is quiet for most modern genres. Pop, hip-hop, EDM, and rock releases are significantly louder. A -14 LUFS master can feel less energetic and less polished on genre playlists.
Mastering to a number ignores the song. If slamming a limiter to -8 LUFS destroys transients, that's a bad master. If a classical piece naturally sits at -18 LUFS and sounds gorgeous, pushing it to -14 only adds unwanted compression.
The "penalty" is just volume reduction. Platforms don't deprioritize or degrade loud masters. They simply turn playback volume down. Audio quality isn't affected beyond the encoding that happens to every track regardless.
So why do some engineers still recommend -14? Because a well-mastered -14 LUFS track with great dynamic range can sound more open and punchy than an over-compressed track turned down to the same level. Valid — but only when comparing a well-mastered quiet track against a poorly-mastered loud one.
What Loud Masters Sacrifice
When pushing louder — from -14 to -10 to -8 to -6 LUFS — limiters and clippers reduce dynamic range. The louder you go, the more you trade:
- Transient impact. Snare snap, guitar pluck, vocal consonants — first casualties of heavy limiting. At extreme levels, everything feels flat.
- Stereo width and depth. Heavy limiting can collapse the stereo image and reduce front-to-back depth.
- Low-end clarity. Bass transients get squashed, and the low end becomes muddy or loses punch.
- Listener fatigue. No dynamic variation tires the ear faster — no contrast between loud and quiet moments.
The trade-off isn't always negative. For genres built on relentless energy — EDM, trap, aggressive hip-hop — dynamic compression is part of the aesthetic. The question is how much serves the music before it starts hurting it.
Genre Loudness Guidelines
| Genre | Typical Range | Notes |
|---|---|---|
| Electronic / EDM | -5 to -8 LUFS | Sustained energy expected. Synthesized material tolerates aggressive limiting better than acoustic. |
| Hip-Hop / Rap | -7 to -10 LUFS | Needs room for vocal clarity and 808 punch. Careful limiting required. |
| Pop | -7 to -9 LUFS | Most major-label releases land at -7 to -8. Consistent energy levels suit limited dynamics. |
| Rock / Metal | -7 to -10 LUFS | Wide spread by subgenre. Preserve drum punch and guitar grit. |
| Jazz / Classical / Acoustic | -14 to -20 LUFS | Dynamics are the music. Don't compress these to compete. |
The Practical Approach
Rather than fixating on a LUFS number, follow this process:
1. Master for the song first. Set your processing to make the track sound its best without watching a meter. Then check LUFS.
2. Compare against your genre. Pull up 3-5 well-mastered reference tracks. Check their integrated LUFS with a tool like Youlean Loudness Meter. Get in the same ballpark.
3. Keep true peak below -1.0 dBTP. Non-negotiable. Going above risks distortion when encoded to AAC, Ogg Vorbis, or other lossy formats.
4. Check short-term loudness in your loudest sections. If your chorus hits -5 short-term while your verse sits at -14, the loudest sections may distort or fatigue. Aim for no more than 6-8 dB spread between quietest and loudest short-term readings.
5. Don't make separate masters for each platform. Platforms have largely converged on similar targets. One well-mastered file works across all of them.
6. A/B test at matched loudness. Louder always seems better (psychoacoustic bias). Turn both tracks to the same LUFS, then judge quality. If your master holds up, you're in good shape.
Common Mistakes
- Mastering everything to -14 LUFS. For EDM, hip-hop, or pop, this will sound weak next to commercial releases. Match your genre.
- Chasing numbers instead of listening. If your master sounds great at -11, don't push to -8 just because a reference measures -8.
- Ignoring true peak. You can nail your LUFS target and still distort on streaming platforms if true peaks exceed -1.0 dBTP.
- Over-limiting to compete. If transients disappear and the track sounds like it's gasping for air, back off until the music breathes.
- Not previewing normalization. Most metering plugins can simulate loudness normalization. Listen at the platform's target level before bouncing your final file.