/posts
25 Oct 2025

File-based music culture

I never hopped onto the streaming music train (Spotify et al.). I'm not sure what kept me off of it originally, but nowadays my motivation is political: I'm not willing to put the access to culture that is close to my heart in the hands of other people with the power to revoke it at any second - out of whim, rightsholder politics, technical fault, bankruptcy, or whatever else. As we know by now, the largest streaming services don't even accomplish what should be their core tenet: Artists need to be financially supported. They aren't. Nobody but the biggest acts and players profit from streaming revenues. I would add an ecological perspective on the problems of streaming services as well, but this shall not be a pamphlet against streaming services, so for the purposes of this introduction I'll wrap it up as it is my reality now: Streaming services are dead to me, long live file-based music culture.

Past to present

I always took my habits around music stored on my own computer as a given, not worth much thought or reflection. As of recent years my perspective has shifted, and I have very conciously re-examined these habits, making obversations, decisions and changes which I will share in this text - not for one particular or political reason (although the politicial is part of this post), but mostly to celebrate and disseminate this culture in and of itself.

Format considerations

My music collection consists almost exclusively of MP3 files, and where I had any say or hand in it, encoded with VBR (Variable Bit Rate) at a bitrate of 220-260 kbit/s (that is -V 0 with libmp3lame). This achieves a few requirements I like: It sounds pretty much lossless to me. It plays practically everywhere. It is more economical than CBR (Constant Bit Rate) encoding and if there's really ever any audible difference, I couldn't care less – I listen to music for enjoyment, not to A/B test-listen for encoding edge cases. Since a while now MP3 is also patent-unencumbered - it belongs to us all, without restrictions.

Recent additions to my collection have also been using the (as far as lossy encoding goes) unmatched Opus format, and I will probably be adding more of it to the collection. On the go I listen to my music collection on a decommisioned, post-EOL smartphone that does not support Opus files, therefore I wrote a script that automatically transcodes these files to MP3 when I sync the library to the device. I also maintain and openly share this script at codeberg.org/simonrepp/recodesync and it can be used with other formats just as well (it is not hardcoded to Opus). If the smartphone should eventually decommision itself for good I plan to scavenge some old-school MP3 player from a local thrift online platform (society is drowning in its own trash and I consider device re-use paramount, also these things have soul).

Keeping only what I love

Some years back my entire music collection weighed in at about 90GB, which was (and possibly will forever remain) its peak size. When I started gathering digital music and internet connections were still slow, downloading a single track from a band occasionally already was an achievement of its own - and often it was hard to obtain complete albums at all. Hence for well over a decade my music collection was mostly a patchwork of songs from different albums.

Sometime around university I began to feel embarrased that my music library looked and sounded like a haphazard download streak from 90s/2000s file sharing platforms (which it was), because back then (being immersed in an academic arts context) I developed a belief that the "true" form of music enjoyment would lie in appreciating the artist's intent to the letter, by listening to albums in full, in order and with dedication.

So in consequence, I got all the albums to which I previously only had a random selection of tracks in full. I also got entire discographies of some underground DIY Post Hardcore act from Tokyo or that sole, genre-defining album released by some historically relevant Screamo act from the US midwest which only ever played a handful of gigs before, because it felt culturally relevant and somehow awe-inspiring to me for a few years. Somewhere along the way I also got into soundtracks a lot and often my impulse would be to get entire soundtracks after watching a movie, and my library grew and grew.

Now at the time of writing my music collection weighs in at a greatly reduced 36.4GB (recall it was 90GB at its peak), and as you might have anticipated, this is mostly due to a rollback of things mentioned above: It turned out I often didn't listen to albums in full, because in the end I realized I didn't actually enjoy many of the tracks on there (while still respecting their artistic value no less), and that there's only so many days in life where I'm musing "I feel like listening to that bootleg of [random underground act] at [obscure club] in [city I never was] from [year where I was still in kindergarten]" and that actually only one minute in one track of that soundtrack album to a movie that I saw contains material I enjoy listening to, while the rest is nerve-wrecking atonal action-cues that make me uncomfortable.

So these days, I pull up my library, sometimes hitting the "random play" button, and I find something I like! Maybe not in that moment, but everything in there sparks joy and can happily stay. And for the weird stuff, the atonal, the raw, the conceptual, the pure artist vision, I love going to concerts, movies, etc., where there is (almost) no limit on weirdness for me.

Tagging considerations

I didn't specifically tend to the integrity and quality of my tags for a long time, but two years ago or so I went through my music library and cleaned things up. I currently mostly use EasyTAG (gnome-based, free software) to edit ID3 tags, using a small, arbitrary set of rules I like:

Genre deserves a closer look: I stick to my own classification of genres, meaning I will often edit genre on new files I add to the library, or reclassify existing files when I see a reason to change my own classification. This is not at all because I believe my classification to be more "correct" than that done by others, but for the simple fact that if I'm the one that needs to find stuff in my own library (and I am), then I should also be the one to assign the labels, no matter how absurd from an outside perspective. E.g. if my brain associates some track by The Cinematic Orchestra with Jazz, then that's what I will classify it as, even if all tagging databases on this planet would tag it as Downtempo, Electronic, World, or something else.

Album too deserves a special mention: I listen to a lot of soundtracks, and at some point I decided to remove any mention of that in soundtrack album titles. "Urban Legend Original Motion Picture Soundtrack" becomes "Urban Legend" and "Edward Scissorhands Original Score (Deluxe Edition)" becomes "Edward Scissorhands". It looks cleaner and is easier to scan for me - I only look for the film titles anyway.

I also made an effort to get (and sometimes hand-select and hand-edit) cover images for everything in my library, usually sticking to a cover.jpg file convention with a 720×720 resolution (or thereabout) - as with audio encoding, good enough is perfect to me in this case, and I wouldn't know what I need more (resolution, quality) for.

Conclusion

File-based music culture is where it's at for me. If I buy as directly as possible from the artists I can greatly support them - streaming subscriptions don't do that (unless one thinks Taylor Swift, Metallica and the military industrial complex need more pocket money). I hope my personal habits outlined here serve as inspiration: Look into your own habits, enjoy what you find. Remake your practice if you're itching for it. Or leave things as they are and bathe in delight how awesome your music collection is! °˖✧◝(⁰▿⁰)◜✧˖°