Every Noise at Once is a scatter plot of musical genres, each positioned by its sonic character — bass-heavy genres cluster toward the bottom, mechanical genres toward the right, organic ones toward the left. There are over 6,000 genres labeled. You can click any of them to hear a representative thirty-second sample. You can click the arrow next to any genre to see its constituent artists, or the angle bracket to hear a playlist.
It was built by Glenn McDonald, who worked at Spotify as what he called "data alchemist." He built it originally as a personal project for his own music discovery, then made it public. Spotify absorbed it unofficially, then officially. It remains the best algorithmic description of the shape of recorded music ever made.
What the site actually is
The visual metaphor — a scatter plot where proximity means sonic similarity — is more accurate than most attempts to map music. Genres that feel similar to listeners actually cluster together. There's a clear gravitational center around the most-streamed mainstream pop, and the further you get from the center, the more specific and unusual the genres become. The far corners of the map contain things like "shoegazing" and "lowercase" and "isolationism" — genres that aren't common words outside of music criticism.
This is a reference you return to rather than consume once. The genre map updates when Spotify identifies enough of a new genre to categorize it, which means the frontier of the map is always slightly different from when you last checked.
The ecosystem around it
The tools that have grown up around Every Noise are worth exploring separately. Several third-party sites use its genre data as a foundation for music recommendations that are considerably more specific than Spotify's own interface allows. One tool lets you construct a playlist from multiple genre coordinates simultaneously, treating the map as literal geography. Another visualizes how your listening history maps onto the scatter plot over time.
Taken together, these form a small alternative to algorithmic recommendation — one built around legible structure rather than opaque suggestion. If you've ever felt like you understood your own musical taste but couldn't describe it precisely, spending time here is instructive.