I joined this group in last.fm after frequently read comments like ‘vote for merge with < Japanese version of artist’s name>’ on many Japanese artists page. It didn’t take long for me to realize the problems not addressed in the current last.fm scrobbling mechanism. These problems mainly plague Japanese artists, especially from anime-music scene.
To sum them up, here’s the problems
- Duplicate pages for artists and tracks, with the original (and official) artist names written in kanji (Sino-Japanese characters), and many versions of Latin (romaji) transliteration. Even a Japanese artist with name ＥＬＩＳＡ is mistaken for an Italian artist with name Elisa (the former being written in ‘wide-latin’ all-caps characters).
- Stats for tracks sung by fictional characters (a character image song or character cover version of a track) are split between the character (with all sorts of romaji transliterations) and the voice actor or singer who really sung them (also with all sorts of transliteration). In this matter I’ll conform to the traditions of western classical opera – identifying the artist as the real-life persona who portrays the character, not the character himself/herself.
- Tracks sung by multiple artists (including fictional characters too), I believe this problem also exists for Western artists too.
The only solution available for these problems is requesting suggestion for correction in each artist’s or track’s page, but this doesn’t work satisfactorily as last.fm site maintainers wait for enough number of users to file the same correction request before actually performing the correction, and it would take many months before a correction actually takes place at best, or maybe it won’t happen at all.
To ask all user-listeners to tidy up their music file metadata is really implausible, since many casual listeners simply either don’t care or don’t have enough time to clean up their metadata, and maybe some of them wouldn’t like to read the track description in characters they can’t read (it’s possible that a small fraction of users don’t have the proper foreign character rendering support in their system). The statistic collection system for artist/performer also has flaws inherited from the CDDB system: it doesn’t distinguish between an artist and a composer (a relevant matter in classical music), and it doesn’t handle multiple-artist identification.
For those problems, I’ve came up with two suggestion. The first is enforcing metadata conformity standard, much like the RFC-ish standard prescribed by Uberstandard.org which is likely to be expanded into digital music encoding standard the same way as Uberstandard has been (Uberstandard is such a Windows-exclusive standard, because although a Linux user can apply it, it relies on a Windows-only proprietary software and a proprietary format noted for its technical limitations; it pissing me off since I use Linux exclusively). The second, much more ambitious one, is to draft a new stats management for last.fm or other scrobbler sites, with open possibility of modifying the current audioscrobbler’s protocol as well. This will include new features such as stats-linking between related artist entities (multiple artists, fictional characters portrayed by a real-life artists), an option for users whether to display artist or track names in latinized version or in the original writings, differentiation between composer and performer, possibility to record additional track’s information – those recommended Vorbis metadata fields, and many others.
While it is unlikely that I would have enough skills to realize the latter, I will propose the first to last.fm community of J-music listeners, although it’s clearly impossible to make all users adopt this standard. Maybe I’ll also try my luck on proposing the second idea to enthusiastic fellows.
First thing first, I’ve got to mend my messy life rhythm. For almost a week I’ve been waking up late, neglecting my heap of laundry, skipping several seminars, not to mention overspending in luxury drinks (namely, Tetra Pak milk) and cigarettes. I’ll definitely be doomed if things keep going on like this.