Asger Risom
For decades, the recording industry engaged in the “loudness war,” using digital limiters to flatten dynamic range so albums would sound as loud as possible on CDs and radio. That arms race degraded audio quality and exhausted listeners’ ears until streaming platforms introduced automatic loudness normalization.
Shuffle a playlist that jumps from a 1980s pop classic to a rock track from 2008, and the sudden volume spike will practically blow your headphones off. For nearly thirty years, commercial record labels operated under the paranoid belief that whoever had the loudest record on the dial won the audience. That obsession sparked an escalating production war that flattened dynamic range, fatigued listeners, and only stopped when streaming platforms quietly changed how playback works.
The race for the volume knob
The impulse to make records louder than the competition started decades before digital software even existed. During the mid-century jukebox era, vinyl 45s pressed with deeper, hotter grooves physically shook the turntable needle harder, grabbing the attention of everyone in a noisy bar. But vinyl had hard physical boundaries: cut a groove too wide, and the stylus would literally bounce out of the spiral trench.
Compact discs tossed those mechanical boundaries out the window when they took over the market in the 1980s. A CD player reads microscopic binary pits with an optical laser, meaning there is no physical needle to dislodge no matter how hot the signal gets. For the first few years of the digital era, mastering engineers used that massive dynamic headroom to produce open, spacious recordings with plenty of room to breathe.
By the mid-1990s, label executives started obsessing over how their songs sounded on commercial radio and multi-disc car stereos. If a new single followed an older track and felt slightly quieter, executives panicked that casual listeners would judge it as weak or poorly produced. The goal stopped being faithful acoustic reproduction and turned into a desperate contest to push peak volume as high as possible. This was not just an anecdotal industry complaint; a 2012 data analysis published in Scientific Reports examined nearly half a million songs released over a fifty-year period and confirmed a continuous, mathematical climb in intrinsic loudness driving the entire era.
Turning waveforms into solid bricks
Digital audio has a hard physical ceiling set at zero decibels full scale. Try to force a sound wave beyond that threshold, and the waveform cannot grow any taller; it simply flattens out, introducing harsh digital distortion. To make a track sound louder without immediate buzzing, engineers had to stop boosting the peaks and start raising the quiet passages instead.
The industry solved that problem with digital lookahead limiters, which act like ultra-fast automated faders clamping down on sudden transients before they can distort. By squashing the loudest momentary peaks—like the sharp crack of a snare drum—engineers created space to drag the rest of the song upward. The average energy level of commercial releases climbed higher with each passing year, steadily eating away at dynamic range.
Open a pop or rock master from the late 2000s in an editing program, and the waveform no longer resembles an undulating sound wave with crests and troughs. Instead, it looks like a solid, impenetrable rectangle of digital noise, a visual pattern that earned the process the name “brickwalling.” Releases like Metallica’s 2008 album Death Magnetic became textbook examples, pushed so hard into digital limiters that audible fuzz and crunch crackled across the final tracks.
The physical toll on human ears
Human ears do not measure emotional excitement strictly by sheer volume; they measure it by contrast. A quiet, intimate verse makes an explosive chorus hit harder because the sudden jump in acoustic energy catches your attention. When an entire arrangement is pinned against the digital ceiling from the first second to the last, that dynamic contrast completely disappears.
The loss of dynamic breathing room also takes a measurable physical toll on listeners. Because the sound provides zero quiet moments for the inner ear’s stapedius muscle to relax, listening fatigue sets in after only a few tracks. Fans often found themselves turning down albums they genuinely enjoyed, completely unaware that unrelenting acoustic compression was wearing out their ears.
Drums and percussion took the worst beating in the loudness arms race. A natural drum strike is a fast burst of transient energy lasting just a few milliseconds, giving rhythm its punch, depth, and physical presence. Flattening those peaks turned snare hits into dull cardboard thuds and smeared bright cymbal crashes into an undifferentiated wash of white noise.
How algorithms broke the trick
The commercial incentive to smash audio files finally collapsed when streaming platforms took over daily listening habits. When Spotify lowered its target playback level to -14 LUFS, as reported by Vice, it removed the perceived advantage that hyper-compressed masters had enjoyed on CD players. Instead of letting every file blast out at its own recorded volume, streaming algorithms began analyzing tracks and leveling the playback field automatically.
That volume normalization instantly flipped the math on brickwalled masters. When the algorithm turns down a slammed, over-compressed song to match a dynamic track, the hyper-compressed song does not sound louder anymore; it just sounds flat, cramped, and thin. Meanwhile, an open mix with punchy transients retains its punch and clarity at the exact same playback level.
In an industry retrospective for Sound On Sound, mastering engineer Ian Shepherd noted that with the vast majority of consumers listening to normalized streams, trying to out-shout the competition is completely self-defeating. When algorithms penalize excessive loudness by pulling down the master fader, dynamic range stops being a liability and becomes an advantage again. Decades of acoustic escalation were dismantled not by audiophile protests, but by a few lines of playback code.
