The spectrograms don’t lie
Nine chord changes. Fourteen timbres. 18 dB of dynamic range. Superstition,
1972.
Morgan Wallen, Last Night,
2023. Number one, sixteen weeks. Three chords. 6 dB. The vocal compressed to a 3 dB window.
Oh come on. This is not taste. This is measurement. The Serrà study: 464,411 recordings, statistically significant declines in diversity across five decades. Average distinct chords in a Billboard number one: 5.2 in the ’70s, 2.8 in the 2020s. Identifiable timbres: 7.4 to 3.1. Rick Beato has been showing people the spectrograms for years. The pictures look like two different art forms.
When Metallica’s Death Magnetic shipped with audible distortion baked into the waveform, the Guitar Hero version sounded better than the commercial release. A video game had higher audio standards than the record industry.
The cafeteria
The discovery optimists say the charts are not the music. Great. The question is about popular music — the music that reaches the most ears, shapes what a generation considers normal. Pointing at the farmers’ market does not change what is in the school cafeteria.
The structural analysts are right about the mechanism. Explanation is not absolution. When every incentive rewards simplicity, the system has produced decline. Calling it optimization
is just decline in a better suit.
Where we concede ground: The 1970s also produced mountains of formulaic dreck nobody remembers. We are comparing a curated archive to an unfiltered feed.
What would change our mind: If the top 10,000 most-streamed tracks returned to 1970s diversity levels by 2035 using the same Serrà metrics.
Read the full synthesis: Is popular music getting worse?