By SongTools.ai Editorial Team
Every major streaming platform normalises audio to a target loudness level. If your master is louder than that target, it gets turned down. Here's how to work with the system.
For most of the 2000s and early 2010s, the music industry was engaged in a loudness war — a race to make masters as loud as possible on the assumption that louder sounds better. The results were releases that were dynamically crushed, fatiguing to listen to, and technically compromised in the process of chasing maximum RMS levels.
Streaming ended the war. Not because the platforms said anything particularly persuasive, but because they implemented loudness normalisation — a system that automatically adjusts playback volume to a consistent target level. When every track is played at roughly the same loudness, the competitive incentive for excessive compression disappears.
Understanding how normalisation works, and what it means for your masters, is now a basic requirement for releasing music professionally.
Loudness in streaming is measured in LUFS — Loudness Units Full Scale. It's a perceptual measurement that accounts for how human hearing weights different frequencies over time, rather than just measuring peak levels.
The major streaming platforms' current normalisation targets:
If your integrated loudness comes in above a platform's target — say you deliver a master at -8 LUFS to Spotify — the platform applies gain reduction to bring it down to -14 LUFS. Your master is turned down by 6 dB at playback.
Here's the problem: when you heavily compress and limit a mix to achieve -8 LUFS, you've traded dynamic range for loudness. After the platform turns it back down, you have a master that sounds compressed and flat at the same playback level as a -14 LUFS master that sounds open and dynamic.
You didn't win anything. You just made your music sound worse.
The practical approach for most genres is to aim for an integrated loudness of between -14 and -13 LUFS, with a true peak maximum of -1 dBTP. Some engineers prefer -0.3 dBTP to provide additional headroom for lossy encoding artefacts.
At -14 LUFS, Spotify plays your track at full volume without gain reduction. Your mix retains the dynamic range that makes it breathe.
Some genres — notably heavily compressed EDM and hip-hop — have conventions that push toward -10 or -11 LUFS. In these cases, you're making a deliberate creative choice to trade some dynamic range for density. That's legitimate, but you're still getting turned down on most platforms.
The most efficient time to check your integrated loudness is at the mix stage, before you send to mastering. A loudness analysis on your mix tells you where your starting point is and helps you brief the mastering engineer accurately.
SongTools.ai's Loudness Analysis tool measures your track's integrated LUFS, short-term loudness, and true peak levels across all major platform targets. Running this check before mastering — not after — means you arrive at mastering with accurate information rather than guesswork.
Peak level as displayed by your DAW and true peak level are not the same thing. When digital audio is converted to analogue or encoded into a lossy format like MP3 or AAC, the samples between sample points can create peaks that exceed the displayed maximum — these are called intersample peaks.
A mix that shows -0.5 dBFS peak in your DAW may have true peaks exceeding 0 dBTP after encoding. This is why the -1 dBTP standard exists — it provides headroom for these intersample peaks to resolve without clipping in the final delivery chain.
Run a true peak check on your master before delivery. If you're over -1 dBTP, you need to bring the master down.
All the tools discussed in this article are available on SongTools.ai — free to explore.