ListenBrainz Music Neighborhood

The ListenBrainz Music Neighbourhood feature is now live!

Use this arcane technology to find your favourite artist and explore their related artists. Watch the pretty colours. Listen to the music. Race your friends from Napalm Death to Britney Spears. Increase the web size and gasp as the artists jiggle into each other. Or simplify your life by exploring a web size of 0.

Don’t wait! No ListenBrainz account required, try it now: https://listenbrainz.org/explore/music-neighborhood/

A cropped screenshot showing colourful detail from the ListenBrainz Music Neighbourhood feature. In the middle is Red Hot Chilli Peppers, connected with lines to a variety of other artists.
Does this say more about the Red Hot Chili Peppers, or their listeners? (hint: the listeners)

Music Neighbourhood might sound new, but it’s been in the works since August 2023, going through a few names. Started by Arshdeep Singh as a GSoC project (project blog post), under mayhem’s mentorship, it was then known as the ‘Artist similarity graph’ . ListenBrainz developers ansh and monkey continued work on it after the end of GSoC, slowly improving it for a number of months, and calling it the ‘Music Web’.

In brief, Music Neighbourhood is a visualisation that connects artists with a high ‘similarity score’, allowing you to dynamically move from artist to artist, exploring their connections. Similarity score is a system that uses listener data to find similar artists, which has already seen use in our Fresh Releases and the LB Radio beta. Artists nearer the center of the visualisation have higher similarity score compared to those on the outside, allowing users to explore near or far. As with a lot of our systems involving listener data, we expect it to increase in accuracy as we get more listeners and listens – if you are getting disappointing results for your favourite artists, it’s time to invite more of their fans to LB! If that fails, you have our permission to sadly shake your head at their listening tastes.

Music Neighbourhood also makes use of another existing ListenBrainz feature, Huesound. Implementing this made a few of the devs involved want to wring the designers neck – sorry! – so make sure to appreciate it. The visualisation pulls the Huesound colours from the top album cover and the top track cover, and uses them in a background and foreground colour combination that hopefully rarely looks like shit (both figuratively and literally), with varied success… it’s safe to say that we could, and possibly will, tweak the colour settings forever, so you may see further changes.

Not just colour tweaks, we expect to to be making more changes and improvements to Music Neighbourhood, so keep an eye out.

As with all of our Explore features, Music Neighbourhood is great on its own, but it also opens a lot of doors. Hopefully we will see this integrated with other ListenBrainz features, perhaps displayed on LB artist or release pages, maybe even expanded into albums, song or genres. We are also excited for others to draw on this for their own external websites, programs and apps, and hopefully see it hacked into something new!

Let us know what you think, and of course share your Napalm Death to Britney Spears high score (27 seconds is your time to beat!)

3 thoughts on “ListenBrainz Music Neighborhood”

  1. I tried it out and it was really great to see it working. Congratulations to everyone involved in the project. It was a privilege for me for to be part of it.

    P.S.- I will make everyone I know to try it out

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.