Finally, after taking too long again (but still maintaining the trend of reducing the time between episodes…), just a few minutes ago I finished tweaking and uploading my most recent topic: “Audio Metadata in Ogg, MP3, and others“.
It’s about 45 minutes long, covering mostly id3v2.3 (MP3 metadata) and vorbiscomments (virtually-everyone-else-that-matters metadata), but I also talk a little about matroska/webm metadata, mp4 metadata, wav, and “windows media”.
This one took a little extra time, because I wanted the episode files themselves to be usable as examples, so I loaded them up with lots of (standards-compliant) metadata, including chapter tags for vorbiscomment-using-files (the .ogg, .spx, and .opus files), multiple attached pictures, geolocation tags, and various other additional tags beyond the small set of basic id3v1 tag names that HPR normally limits itself to. That way, after you listen to the files, you can play with them. It’s a podcast, AND an educational toy! Doing this meant doing the encoding for the final files myself, since that was a heck of a lot more reasonable than trying to talk the HPR admins into going through and putting all my special metadata in for me during processing. (That also means if there IS something wrong with any of the metadata OR the audio encoding, it’s all MY fault and not HPR’s).
Once the files are up at hackerpublicradio.org, I’ll put in a link to the opus version for awesome, special people who prefer opus and/or don’t want to wait for the episode to come up in the HPR queue (though you can also listen to the ogg or speex versions (or even mp3 if you must) from hackerpublicradio.org directly as soon as they’re set up, even before they show up in the daily queue).
As mentioned in the episode, I’m trying to put together recommendations for mapping tagnames between vorbiscomments and id3 (and maybe eventually matroska/webm as well). I’ll be spending some time putting some of that online tonight. I’ll try to keep the page updated as events warrant.
The next episode should come very soon, and will be a short (no, really this time) episode on a problem I had years ago on BTRFS and how I solved it. I’m intentionally NOT trying to turn it into a general episode about btrfs, because otherwise I’ll be researching for another month or two and then ending up producing another hour-long extravaganza. I’m aiming for just 10-20 minutes on this one, I swear.
After that, I’ll decide. You’re welcome and encouraged to take a look at my “Potentially Upcoming Topics page (regularly updated) and comment on any subjects that you’d like to hear me attempt.