Series
Homelab Music: Discovery on a Library You Own
Building Spotify-style music discovery at home: the signals and trade-offs, track-aware scrobbling, sending every listen to ListenBrainz, sonic analysis with AudioMuse-AI, and using AI to make recommendations from your own history.
5 parts · first published
- 01
The Homelab Music Recommendation Field Guide: Every Signal, Trade-off, and Winner
A comprehensive reference for building Spotify-like music discovery at home: identity, metadata, sonic analysis, loudness, behaviour, collaboration, playlists, smart fades, household discovery, and how to choose between competing approaches.
·22 min read - 02
From YouTube to a Scrobble: The Case for Track-Aware Listening
YouTube is a major source of music listening, but ordinary scrobblers treat an album upload or DJ mix as one video. This article maps the ecosystem, explains the splitting problem, and outlines what an ideal open-source solution could look like.
·17 min read - 03
Scrobbling Everything to ListenBrainz: Navidrome, an iPod Running Rockbox, and Beyond
A recommender is only as good as the history behind it. This is how every play in my homelab, from Navidrome on the server to an offline iPod running Rockbox, ends up as a clean ListenBrainz listen, and how that history comes back as playlists.
·9 min read - 04
AudioMuse-AI: Sonic-Analysis Recommendations on My Own Library
How AudioMuse-AI analyses audio, how it runs in my homelab beside Navidrome with a local LLM on a separate GPU host, what its playlists are good and bad at compared with ListenBrainz, and how I clean up the playlists it generates.
·8 min read - 05
Using AI to Build My Own Music Recommendations from ListenBrainz Listens
Combining my ListenBrainz history (iPod plays included), AudioMuse-AI sonic similarity and a local LLM into playlists that explain themselves. What already runs in my homelab, what is only a design, and how to tell whether any of it works.
·7 min read