A new browser tool lets Spotify Premium users fool around with the music streaming platform’s famous recommendation algorithm. Arielle Vaniderstine, an engineer at the company, posted a link to Glitch on Monday of a beta build of the tool.
Called “Nelson,” the tool displays several sliders representing the “audio features” that inform the algorithm. From there, users can play around with the sliders, manually altering the inputs.
Acousticness: A confidence measure from 0.0 to 1.0 of whether the track is acoustic. 1.0 represents high confidence the track is acoustic.
Danceability: Describes how suitable a track is for dancing based on a combination of musical elements including tempo, rhythm stability, beat strength, and overall regularity. A value of 0.0 is least danceable and 1.0 is most danceable.
Energy: Energy is a measure from 0.0 to 1.0 and represents a perceptual measure of intensity and activity. Typically, energetic tracks feel fast, loud, and noisy.
Instrumentalness: Predicts whether a track contains no vocals. “Ooh” and “aah” sounds are treated as instrumental in this context. Rap or spoken word tracks are clearly “vocal.” The closer the instrumentalness value is to 1.0, the greater likelihood the track contains no vocal content. Values above 0.5 are intended to represent instrumental tracks, but confidence is higher as the value approaches 1.0.
Liveness: Detects the presence of an audience in the recording. Higher liveness values represent an increased probability that the track was performed live. A value above 0.8 provides strong likelihood that the track is live.
Speechiness: Speechiness detects the presence of spoken words in a track. The more exclusively speech-like the recording (e.g. talk show, audio book, poetry), the closer to 1.0 the attribute value. Values above 0.66 describe tracks that are probably made entirely of spoken words. Values between 0.33 and 0.66 describe tracks that may contain both music and speech, either in sections or layered, including such cases as rap music. Values below 0.33 most likely represent music and other non-speech-like tracks.
Tempo: The overall estimated tempo of a track in beats per minute (BPM). In musical terminology, tempo is the speed or pace of a given piece and derives directly from the average beat duration.
Valence: A measure from 0.0 to 1.0 describing the musical positiveness conveyed by a track. Tracks with high valence sound more positive (e.g. happy, cheerful, euphoric), while tracks with low valence sound more negative (e.g. sad, depressed, angry).
It might seem complex, but just remember that musical taste is itself a sort of equation made up of similar variables. You might not think in terms of “valence” or “energy,” but, broadly speaking, when you’re getting a feel for a new song, you’re probably paying attention to whether you can dance to it, and whether it’s uptempo or slower. The browser tool just takes the way Spotify quantifies these elements, then lets you play around with them.
For me, it seemed like a cool way to explore outside my own musical tastes. I listen to a lot of upbeat pop and I like singers with big voices. With Nelson, I can try a different genre, let’s say metal, then ask Spotify to recommend me metal songs that are both fast-paced and heavy on vocals.
For those that want to go further into Spotify’s algorithm, the beta Developer site lets you look up the audio features for individual tracks. “After the Storm” from Kali Uchis is a great, throwback R&B slow jam. Computers don’t know what any of those descriptors mean, so here’s how Spotify quantifies a throwback R&B slow jam:
"danceability": 0.589, "energy": 0.912, "key": 9, "loudness": -3.748, "mode": 0, "speechiness": 0.0472, "acousticness": 0.000245, "instrumentalness": 0.00167, "liveness": 0.571, "valence": 0.579, "tempo": 99.995, "type": "audio_features",
I took the values, fooled around with sliders, and tried to become my own manual recommendation algorithm. Keeping the values the same while switching up the genre made for some surprising recommendations, like Ministry’s “You Know What You Are” and Liam Gallagher’s “Wall of Glass.”
We now live in a world full of algorithmic feeds and recommendation rabbit holes, making Nelson a fun time-waster that’s also a cool look behind the behind the scenes, showing how the digital sausage is made.