Blog Archives
Discoverability
I think about the tragedy of discoverability a lot.
Maybe “tragedy” is not the right word, but stick with me for a moment. Think about your absolute favorite song, or movie, or game, or book. Was encountering that now-favorite thing inevitable? For some things, maybe the answer is Yes. If you had a Super Nintendo in the 90s, odds are pretty high that you were aware of Final Fantasy 3 (6) or Super Metroid or other bangers that were pre-internet famous even in their day. Then again, perhaps that was a bias I had from reading gaming magazines like Nintendo Power or Game Pro back then. And, of course, it presupposes you (or your parents) had the cash to purchase it for you.
When looking at my music collection though, almost all of my favorite songs were never in any way inevitable. Fifteen years ago, I was working an extremely boring data entry job that nevertheless allowed us to listen to music. I had the entire internet at my disposal, and yet you don’t know what you don’t know. Where does one listen to music, that isn’t just the same 17 songs on the radio? My solution was these monthly indie rock playlist drops, such as BIRP, where you can download like a hundred separate songs, and then winnow them down to the usually 2-3 that were my style. From there, I would download that musician’s discography and then winnow that down, finding more diamonds in the rough. Rinse and repeat.
If I did not have that job, or did not discover the indie rock playlist – which I honestly have no recollection of where I would have encountered it – then some of my favorite songs would just… not exist, from my perspective. Yes, I would still have had my prior favorite song that I did encounter, so in a way it’s relative. In a more concrete way though, it’s very much not relative because my new favorite had greater depths and more fulfilling pathos than what came before.
My job has changed quite a bit over the years, and I no longer felt it was a good use of my time sorting through a hundred songs of what felt like UK punk rock for a single good song. So, I stopped around 2018, it looks like. How many diamonds have slipped through my fingers in the last eight years? Maybe fewer than I think, given I have leveraged more Youtube algorithms for exposure to new stuff. Then again, just like with AI generally, it’s exceedingly proficient at giving you more of what you like, rather than something totally different that you might like more.
I do not know if there is a solution out there for discoverability, other than just putting yourself out there. The corporate machines of the world are certainly spending billions of dollars in trying to put your eyeballs on their products, but for the average piece of content, it’s likely random happenstance at best. And that’s really too bad.
In unrelated news, I’ve signed up for Blaugust again this year, and threw my hat in over at Warp Point.
In An Age 
Interlude
May 16
Posted by Azuriel
When it comes to music, I have a System.
Stage 1 is Acquisition. My ability to discover new music is very limited, as I did not have much of a commute even pre-pandemic for radio, and I typically listen to NPR besides. Therefore, I pretty much only hear snippets of songs from random Youtube clips, memes, or Indie-Rock playlists. Once I do hear something I like, I immediately try to locate the entire discography of the group “just in case” there are other songs I may enjoy. The amount of times this has worked out for me is small in the aggregate, but when it has worked out, I ended up finding some of my favorite songs of all time, so it’s worth it.
Stage 2 is Sifting. I have a dynamic playlist called Unrated that will display every song in my library that does not have a 1-5 Star rating. Once a rating is decided, it drops from the playlist and I continue down the list until my queue is empty. Roughly speaking, ratings mean:
To give an idea about distribution, I have 3270 songs in the 3-Star category (80%), 721 in 4-Star (18%), and just 100 in 5-Star (2%).
Stage 3 is Assignment. Once everything has a Star rating, it goes into my Group Work dynamic playlist. It is here that I assign it a “Grouping” category. I base these off of colors that correspond to mood:
Once this last piece of metadata is complete, songs that have survived the gauntlet are automatically sorted into dynamic playlists. If I want to chill and do some work? Open the Green playlist and hit Random. Roadtrip? Load up Red and press Play. Want to just feel sad for a while? Blue. This has worked out much better for me than traditional playlists, as sometimes I forgot to add new songs to old playlists. Plus, some songs fit into multiple moods, which I can handle by giving them multiple Grouping labels (“Green, Blue”).
To be honest, I have no idea if this is convoluted or not. My default assumption is… Yes. But I haven’t really heard how other people tackle this sort of thing. Surely everyone has playlists of some kind, yeah? Do people just use Spotify these days? What is your System? Partially because I’m curious, and partially so I can determine what improvements (if any) can be assimilated into my own.
Addendum: Program-wise, I am using MusicBee as the player on PC, and Rocket Player on my phone. I used to use iTunes on PC and synced things easily with Rocket Player, but at some point that method failed and I could never figure out why. So these days I just bulk copy/replace files on the phone periodically.
Acquisition-wise, TPB used to be good enough, but these days I just rip from Youtube.
Posted in Commentary
5 Comments
Tags: Convoluted, First World Pirate Problems, I Have a System, Music, MusicBee, Playlist