Music Is Louder and More Boring Than Ever, Says Math

How hard did that headline blow your mind? Imagine: music really is crappier now; you’re not just turning more snobby. It’s louder, not just because you’re playing it louder because technology is making it much easier and cheaper to give yourself tinnitus, but because modern studio techniques allow music to be produced louder. And it’s duller; the variety of timbres (instrumentation, recording techniques) used over the last 50 years in music has dropped significantly across a wide variety of genres, after peaking in the ‘60s, while we’re basically using the same set of pitches (melodies, specifically) as we were in the ‘50s, while sacrificing the variety of ways in which we combine those melodies. This is all according to a paper out this week in Nature’s open-access journal Scientific Reports, courtesy of Joan Serrà of Barcelona’s Artificial Intelligence Research Institute et al.

This isn’t Top 40 sampling. The research uses Columbia University’s Million Song Dataset, which contains beat-by-beat data on one million different Western songs. They then do basically the opposite of sonification — different musical facets (melodies, instrumentation, etc.) become data points and, then, can be subject to regular old statistical anaysis models, like Zipf’s law, which describes a frequency relationship in which one musical chord’s frequency is inversely proportional to its rank in a frequency table of different chords. (Zipf’s law is usually used for linguistic analysis but works pretty well for music too, it turns out.) That is, the most popular chord is twice as common as the second most popular and three times as common as the third, and so on.

This relationship holds, more or less, for the 50 years of analysis, but what they found was a different distribution of chords within pieces of music. Rather than common chords being seperated by less-common — with interesting and original results — common chords are more and more just seperated by other common chords. The chords are the same as ever, but the combinations are more boring. Meanwhile, their research suggests “a growing homogenization of the global timbral palette,” according to the paper. “It also points towards a progressive tendency to follow more fashionable, mainstream sonorities.” So the ways in which we style and color music have dropped, suggesting a mainstream-as-black hole kind of effect; timbres don’t go away to be recycled in some new way, they disappear past an event horizon.

And finally, Serrà et al found an average 9 percent increase in decibel level over the past 50 years within the dataset. Which itself contributes as well to music becoming more-boring — the more you max out a digital piece of music, the more dynamics you lose. The conclusion from all of this, in the author’s words: “Much of the gathered evidence points towards an important degree of conventionalism, in the sense of blockage or no-evolution, in the creation and production of contemporary music.”

“Blockage,” I like that. It suggests some potential for relief, that if only we as music consumers and creators can figure out exactly what that blockage is and how to remove it. It’s a bunch of things, really, and probably a great many having to do with how we put price tags on music, and a decrease in the industry’s willingness to take risks on innovation, which lately is increasingly driven by desperation. Most likely, it’s the bare fact that music is even driven by an industry. Maybe the blockage will just have to remain, the necrosis of popular music that’s already underway will continue, the tissue will die, and music can only then regenerate.

Please excuse the doom. And finally, for some amount of hope, I’ll direct you over to the Type Records Soundcloud page.

• Music Has a Right To the Blank Tapes of Children
• Music Industry Dinosaurs To Google On Open Internet: Wah!
• The Weirdest Piracy Haven Of Them All: Goodbye For Now, Megaupload