You would think that a legally purchased album of traditional Volksmusik, produced and curated by #BayerischeRundfunk and played by their chamber orchestra (with more acoustic/traditional instruments) and then sold through Amazon (who have had presence in .de for decades) would *not* have *standard* German characters in the filenames all mangled?! (yet the rest of the MP3 metadata is correct!) #unicode #encoding #fail
@vfrmedia Double-encoding. Probably a borked conversion from some ISO encoding to UTF-8 (or any number of such steps, I can't find a matching one right now)
@elomatreb that could well explain it - I was in late 1990s configuring broadcast automation equipment that BR bought from UK and were using in then new studio complex (they were ahead about going digital compared to UK BBC etc).
I suspect this music comes straight out of what would be their music library (its linked to a radio/TV series) and might have had the older encoding in the titles since the 1990s..
@vfrmedia No need to go back as far as the 90s, ISO8859(-15) was a database encoding default for *a long* time, and tends to stick around even today in enterprise settings Germany (as transcoding a database is dangerous and/or a lot of work).
@elomatreb ISTR our equipment just about handled ISO 8 bit character sets (would have choked on Unicode) and the other broadcast automation kit it interfaced with was similarly old (and one big German manufacturer of it went bust in the 2000s). To be fair BR (and DE public broadcasters as a whole) have done an amazing job of making their new and historical content available online (way better than UK) and don't even geoblock so I can understand if they need to keep legacy databases running..
@vfrmedia Normally there should be no need for actually keeping data in legacy encodings. UTF-8 can store *everything*, unless you made private modifications but even those can be worked around with the Private Use Area.
It just requires some know how and an initial time investment. For equipment that can't handle new encodings the "good" solution is to transcode and output time, and store in a modern encoding.
@vfrmedia *transcode at output time
@vfrmedia absolutely nobody does metadata correctly. It's very frustrating.
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!