If you’re like me, you have thousands of songs in your library and only dozens that you listen to regularly. This a strange problem and growing storage capacities promise to make it worse. While the hand-curated playlist remains the pinnacle of listening experience craftsmanship, for the casual music lover it’s easy to forget songs that you love and it’s often too slow and frustrating to comb through thousands of songs to put together a new mix. Shuffle is too random when you unleash it across an entire library and features like Apple’s Genius aren’t smart enough. Faced with an overwhelming abundance of tunes and inadequate tools to manage them, it’s easier to just retreat to one the few playlists you already have. Enter Gravity Mobile’s Habu.
Habu is a playlist generator designed to create highly customized playlists around highly granular moods. “Habu was created for people who’d rather spend their time discovering new music than creating custom playlists,” their marketing material explains. Habu works by matching your music collection to Gracenote’s database of “mood profiles” for songs. Once it’s done this, you can use the app to create new playlists based on your mood (e.g. “Today I feel like listening to some Abstract Beat like Butterfly by Talvin Singh.”). Moods are mapped to a two dimensional plane, one axis runs between Positive and Dark, the other runs from Calm to Energetic. All told, there are 25 grouped moods and 100 individual moods available.
“We believe introspection is often lost in a world of massive consumption and regurgitation,” says Eric Eisher, Design Lead for Habu, “As designers, we have an awesome opportunity to help people learn more about themselves and others.” This opportunity begins with a data-visualization approach to presenting your library back to you. The key insight of Habu is that once your collection of music gets large enough, you no longer have a curation problem, you have a data management problem.
“How do we take 100 points of entry and make them scannable and engaging?” says Eisher. “That is the question we started with.” All your songs are mapped to those 100 moods, which are laid out along the two axes in the visualizer. If more songs are mapped to a particular mood, a larger more yellow circle is drawn. If there are less songs, the mood gets a smaller, more purple circle. The app creates a mood map of what you’ve got, with circles that get bigger and smaller depending on how much of your collection matches that profile. It allows for a relatively high information density.
Eisher says that the importance of this approach will become more and more apparent as users move from their 12,000-song personal library to a 12,000,000-song streaming service. The goal is to ensure that an enormous amount of data could be made to seem manageable. Eisher says they went through many different design proposals before settling on the bubble map. “Our first iteration of Habu gave each song in a user’s library its own shape and form. On paper, the idea still sounds uber-dreamy, but there were so many caveats that stopped this idea from growing.”
By refocusing the app on the most basic interaction–one-touch playlist creation–Eisher says that it opens up more interesting opportunities for interacting with the data in the future. “We have the potential to help people answer some pretty cool questions,” he says, suggesting that Habu could enable people to find out what music their city or neighborhood was into, or comparing their libraries with that of their friends. “These engagements are where we’d like to see Habu evolve to, but the collection, visualization, and interaction of these data sets needs to be dead-simple to play with, learn from, and digest.”