Loudness, LUFS, and why streaming platforms changed everything about mastering
23/05/2026
When Spotify introduced integrated loudness normalisation back in 2013, most mastering engineers noticed. When Apple Music followed, it became impossible to ignore. The rules of mastering had quietly shifted — and a lot of musicians were still playing by the old ones.
This post explains what LUFS actually means, why streaming platforms use it, and what it means for how your music should be mastered today.
What LUFS means
LUFS stands for Loudness Units relative to Full Scale. It is a standardised measurement of perceived loudness — not peak level, not average level, but something closer to how loud a piece of audio actually sounds to a human ear over time.
The older way of measuring loudness — dBFS (decibels relative to full scale) — only measured the peak amplitude of a signal. A heavily compressed track could have a high peak level but sound much louder than a more dynamic track at the same peak. LUFS accounts for this by integrating loudness across the duration of a programme.
In practical terms: two tracks can have identical peak levels but sound completely different in volume. LUFS is the measurement that gets closer to capturing that difference.
Why the loudness war happened
For years, the goal in mastering was simple — make it as loud as possible. Louder tracks sounded more impressive on radio. They cut through in a club. They grabbed attention in a listening test. Labels and artists pushed for it, engineers delivered it, and the result was a decade of music that was heavily limited, dynamically crushed, and fatiguing to listen to.
This is what people mean when they talk about the loudness war. It was not about making music sound good. It was about making music sound louder than the track before it. Dynamics — the difference between quiet and loud moments — were sacrificed to gain a few extra dBs of perceived loudness.
The technical ceiling is 0 dBFS. Once you hit that, you cannot go louder. So instead of turning things up, mastering engineers used limiters to compress dynamic range and raise the average level without technically exceeding the ceiling. Waveforms that used to look like mountains started to look like bricks.
What normalisation changed
When Spotify introduced loudness normalisation, it began measuring the integrated loudness of every track uploaded to its platform and adjusting playback volume so that all tracks played at roughly the same perceived level. The target is around -14 LUFS.
Apple Music uses -16 LUFS. YouTube uses -14 LUFS. Tidal uses -14 LUFS. The exact number varies slightly by platform, but the principle is the same across all of them: if your track is louder than the target, it gets turned down. If it is quieter, it gets turned up.
This changed the calculation for mastering entirely. If you master a track to -8 LUFS — extremely loud, heavily limited — Spotify will turn it down by 6 dB at playback. All that limiting and compression was done for nothing. Worse, the heavy processing needed to achieve that loudness often costs you clarity and dynamics that you cannot get back.
A well-mastered track at -14 LUFS will play at exactly the same volume as an over-limited track at -8 LUFS on Spotify. But the well-mastered track will sound more open, more dynamic, and usually better.
What a good target actually looks like today
For most streaming releases, a target of -14 LUFS integrated is a sensible starting point. This is not a rigid rule — genre matters, the nature of the music matters, and different platforms have slightly different targets — but it is a reasonable default for most modern releases.
A few things to understand about that number:
It is an integrated measurement. LUFS integrated measures the average loudness across the whole track. A quiet introduction followed by a loud chorus can have the same integrated LUFS as a consistently mid-level track. The measurement accounts for the whole piece, not just the loud parts.
True peak is a separate concern. Even at -14 LUFS, you need to manage true peak levels. The standard recommendation is to keep true peak below -1 dBTP to avoid inter-sample clipping during the encoding process that streaming platforms apply to your audio.
Louder genres need a different approach. A metal track mastered to -14 LUFS may sound underwhelming against genre expectations, even if it technically plays at the right level. Understanding the genre and the context matters more than following a number blindly.
What this means when you send a track for mastering
When you send a mix for mastering, leave headroom. A mix hitting -6 dBFS on the peaks gives a mastering engineer something to work with. A mix that is already clipping, already limited, already pushed to 0 dBFS on the bus — there is very little a mastering engineer can do to improve that. The damage is already in the file.
Remove any limiting or heavy compression from your master bus before you export a mix for mastering. That processing should happen at the mastering stage, applied with full knowledge of the target format and platform.
If you are mastering for vinyl, the target is different again. Vinyl cannot handle extreme bass levels or very high integrated loudness. The requirements are different enough that a vinyl master is typically a separate deliverable from a streaming master — something to discuss before the project starts, not after.
The short version
The loudness war is effectively over on streaming platforms, because loudness normalisation made it pointless. A well-mastered track at -14 LUFS sounds better than an over-limited track at -8 LUFS — and on Spotify, Apple Music, and YouTube, they play at exactly the same volume.
Master for how it sounds, not for how loud it measures. Leave headroom in your mix. Trust the process.
If you want to talk through where your tracks are sitting and what they might need, get in touch.