The loudness war didn’t die. We just stopped fighting it where you could see the corpse.
The loudness war isn’t over. Streaming normalization just turns loud masters down without giving back the dynamics they destroyed, so producers keep crushing tracks anyway.
Here’s the story you got told. Spotify normalizes to -14 LUFS by default (its own artist docs say so). YouTube and Tidal land on -14 too, Amazon Music measures around -14 even though it doesn’t publish a number, and Apple Music’s Sound Check, now on by default, sits at -16. So there’s no point smashing your master into a brick anymore. Loud doesn’t win. The platform turns everybody down to the same level, so the guy who mastered quiet and dynamic sits right next to the guy who clipped his hi-hats into dust, and they hit your ears at the same volume. Peace in our time. Mix engineers can finally breathe.
Cute story. That’s the brochure. Here’s what actually happens.
Normalization is a volume knob, not a time machine
Normalization adjusts one number: integrated LUFS, the average loudness over the whole track. It doesn’t un-crush anything. Slam a master to -6 LUFS and Spotify pulls it to -14 by pulling the fader down. That’s it. Everything you murdered to get loud is still dead, just quieter. The peaks are still glued to the ceiling, the transients are still flattened, the kick still has no punch left in it. You can’t get the dynamics back by lowering the level, the same way you can’t un-burn toast by turning the toaster off.
So why is everything still mastered loud?
Three reasons, and none of them are stupid.
One: defaults. Most people never touch loudness normalization, and it’s off in a depressing number of places: DJ software, raw-file playback, Bluetooth speakers doing their own thing, your friend’s car aux. The moment your track plays somewhere normalization isn’t active, loud wins again, instantly. Master quiet “for the platform” and you get smoked everywhere else.
Two: loud still feels “better.” It’s the louder-sounds-better bias. Louder reads as fuller, more finished, even when it’s objectively worse. Play someone the dynamic version and the crushed version matched for loudness and a lot of them still pick the crushed one because it feels denser. The catch: when researchers level-match and run the test properly, that preference is way shakier than the folklore claims. Moderate compression gets a slight edge at best, heavy compression is a wash or actively worse, and untrained listeners often can’t even tell. Doesn’t matter. The feeling drives the request.
Three, the one mastering engineers say off the record: clients still ask for it. “Make it competitive.” “Make it loud like the reference.” The reference is loud, so the master gets loud, and normalization quietly undoes the loudness while keeping all the damage. Worst of both worlds, paid in full.
What it does to dance music
Now the part I actually care about, because this is a dance music blog. Electronic music is dynamics. The whole language is contrast: the filtered build that drops the bottom out so the bass return hits like a truck, the breakdown that pulls everything back so the room can breathe. A loudness war master flattens that into a single screaming plateau. When the loudest moment and the quietest moment are the same loudness, there is no drop. There’s just more stuff happening at the same level that’s been hammering you for four minutes. The kick stops punching because there’s no headroom around it for the transient to live in. A perfectly engineered drop, reduced to a texture change.
DnB makes it ugliest. Jungle and drum-and-bass run on transient sharpness. A crushed amen break is a dead amen break, all the snap squashed flat. You can hear the difference next to a record mastered by someone who actually trusts dynamics. A lot of the better dub-leaning techno still does. The Basic Channel and Chain Reaction lineage (Moritz von Oswald and Mark Ernestus’s Berlin labels) built its whole sound on space and restraint. Put that next to a brickwalled festival master and notice your body reacts to it differently. You feel the drop instead of just hearing that the arrangement changed.
Here’s the part that should end the argument: normalization actively punishes the crushed master now. On a platform like Spotify that turns quiet tracks up (as far as the peaks allow), a dynamic master that left headroom gets lifted toward the same -14 average and keeps its transients, so it hits harder than the brick sitting next to it. And even on the attenuate-only platforms (Apple Music, YouTube and Tidal only turn loud tracks down and never boost), the loud master is the one getting pulled back toward the dynamic one. Either way the math runs against the brick. The thing producers crush their tracks to win? It loses on the exact platforms they’re crushing them for.
Test it yourself. Pull up a track you love from 2008 and a remaster from the last few years, turn the old one up to match, and listen for the kick. The old one breathes. The new one is a wall.
The loudness war was never really about loud. It was about fear: fear of being the quiet one in the playlist, fear of sounding less finished than the reference, fear of telling a client no. Streaming normalization handed everyone a permission slip to stop fighting, and most of the industry said thanks and kept crushing anyway. The ceasefire got signed. Nobody told the masters.
The smart move is the one almost nobody makes: master for the drop, and let the platform worry about the volume.
Related: Virtual Riot & YDG - Paralyzed · Sustance - Dub Special w/ Mad Sam · Fred V - Mistakes ft. Luxtides