A few discoveries in how to run TDarr.
A single LLM call rarely produces perfect results. Ask an AI to organize a movie library into collections and you'll get inconsistent output—sometimes creative, sometimes missing obvious franchises, sometimes grouping unrelated films under vague themes. The solution isn't a better prompt. It's a structured pipeline that breaks the problem into focused passes, each with its own validation. This post documents a 5-pass analysis pipeline that combines local embeddings with LLM validation. The emb
Organizing a large media library into meaningful collections is harder than it sounds. With 1000+ movies, you need something smarter than keyword matching. "Star Wars: A New Hope" and "Star Wars: The Empire Strikes Back" share a franchise, but a text search for "Star Wars" won't catch "Rogue One: A Star Wars Story" if the title varies. Movies about heists aren't all titled "heist." Thematic grouping requires understanding what movies are about, not just what they're called. The solution is sem
I've spent the last few weeks setting up a Plex server on a Synology NAS drive and I have finally gotten it working reasonably well. I haven't set up the arr suite or whatever yet because I can really only tolerate spending so much of my free time debugging docker containers but I'll get there. This was a long and mostly tedious process, but now I have a self-hosted media library I can access anywhere through Plex. I struggle a lot with making myself do stuff like this. If I'm not working on a
For the sake of full disclosure, I started writing this guide with the intention of making a clearer version of the NASCompare setup guide, but over the course of a few days of work discovered that guide didn't work at all and ended up pretty much directly following the instructions from Dr. Frankenstein's guide, which is thorough and probably does a better job of covering all your bases than this will. I put enough effort into writing this that I figured I'd publish it anyway but if you get los
Every time I purchase digital music it gets downloaded and archived the exact same way.
To track the music I listen to I need to add records for artists and albums associated with them. This is something I've tackled manually — add an artist name, slug, description and a record for each album.
I've talked about building my own music scrobbler , I've talked about improving it and I've complained about wanting to stream my own music and then I wrote a retrospective about it . I've settled on something that works, so let's look at it in a bit more detail.
I think my self-hosted scrobbling implementation is, at this point, fairly stable. The charts I'm calculating track pretty closely with last.fm's and I'm inclined to chalk the difference up to how we're calculating/slicing dates.
I wrote (fairly) recently about implementing my own scrobbler using Plex webhooks, edge functions and blob storage. So far — so far — this has worked quite well.
I've written before about embedding music into my site and I've largely used Last.fm to do so. Their API is rather extensive, though it is showing its age — the default response format is XML, they've dropped artist images and have intermittently failed to return album art. ListenBrainz is great, but client support is still lacking. I've also tried charting Apple Music data from their (quite limited) API.