Tips for discovering new music

[Altruizine]

Sorry if a thread like this already exists.

As a teenager I used to probe the web for new music like crazy for a couple of years but pretty much stopped completely due to some internal time management and motivational conflicts that wouldn't allow me to indulge so much. My taste has been more or less stuck since then with random new discoveries here and there.

I mostly used last.fm "similar artists" but later on other means too (blogs and the like). Forums too, of course.

Nowadays you'd expect there to be tons of innovative and convenient ways for this. I was thinking "what if there's a Tinder for discovering new music that plays snippets and you swipe on them" and lo and behold someone had made that

What do you use?

[Glen_Glog]

Listen to 1990-1994 rap music and work back from the samples -- worked for me

new to you or just new new?

[Altruizine]

new to you or just new new?

New to me but with new new being particularly appetizing in case there's something novel happening. Not contrived novel like early 2010s microgenres but something like, advanced music tech that sounds unlike anything before or clever psychoacoustics or something.

Or is it all still just binaural beats?

[Cupz]

I tend to have soulseek open in one half of my screen and discogs.com in another. I love combining weird sub-genres of music and see what pops up.

[llydia]

I tend to have soulseek open in one half of my screen and discogs.com in another. I love combining weird sub-genres of music and see what pops up.

ur on slsk??? same... yeah i like to find weird stuff on there. Lol I cannot name some shit publicly what I have found there, just some bloody good stuff xD

[Negamuse]

We need to bring personal music blogs back, I found some great stuff from them back in the day. I think there's still a few from the tape scene hanging on.

YouTube playlists? I found a few interesting ones trying to chase down the last few unidentified societas x tracks.

[Sherbet Head]

In my experience...

• Mixes/playlists/DJ sets curated by my favourite artists probably has the highest success rate for discovery.
• Keeping up to date with my favourite independent record labels. If I notice that a few good artists are on the same label, chances are there's more decent stuff in their roster worth checking out.
• Forums and social media.
• Annoyingly, I've found Spotify's discovery algorithm to be fruitful. It throws some garbage my way, but occasionally it turns me onto some absolute gems.

[Negamuse]

Algorithms are tricky, some recommendations are stickier than others. Like on YouTube videos about 3d printing, right wing politics, Lego or trains are complete tar-pits, you can't watch one without it taking over your recommends.

Well same with music, I don't really get post-rock, just not my thing but like obviously so close to a lot of stuff I really DO like. So it's hard to get good Spotify recs that haven't involved human curation, without getting swamped by post-rock acts I'm not into

That said I'll often put on one track and see where the song radio takes me, sometimes it feeds me good things. I've been listening to a lot of mid teens Scandinavian bands I totally missed first time round cos of that

[Alvin]

Listen to 1990-1994 rap music and work back from the samples -- worked for me

Nice. Been listening to a lot of that recently so I should take that advice. Do you use https://www.whosampled.com/ or something else?

[Glen_Glog]

Listen to 1990-1994 rap music and work back from the samples -- worked for me

Nice. Been listening to a lot of that recently so I should take that advice. Do you use https://www.whosampled.com/ or something else?

I use whosampled when I cant remember the name of the original track usually during a conversation or argument -- got any fave albums so far? Would love to hear

[Alvin]

Listen to 1990-1994 rap music and work back from the samples -- worked for me

Nice. Been listening to a lot of that recently so I should take that advice. Do you use https://www.whosampled.com/ or something else?

I use whosampled when I cant remember the name of the original track usually during a conversation or argument -- got any fave albums so far? Would love to hear

Good idea. There’s many I could post but these 3 immediately led to interesting discoveries:
[media]https://youtu.be/xC3U3gIOR4w[/media]
Samples this:
[media]https://youtu.be/tTNb9Ebdzyk[/media]


[media]https://youtu.be/PlhOwhljE5M[/media]
Samples this:
[media]https://youtu.be/v-jMxU0S50Q[/media]


[media]https://youtu.be/vFRiyTaB2A0[/media]
Samples this:
[media]https://youtu.be/D1NLo8siLiQ[/media]

[Alvin]

Your turn.

[Altruizine]

I decided to finally ask ChatGPT about "cutting edge 3d audio" and this was the first recommendation it gave me

https://ex-log.bandcamp.com/album/entanglement

I've never experienced anything like this. This is SO good. Some of the other recommendations were fire too. The best part is that I've only got some 20€ bluetooth headphones because all my higher quality headphones are broken. Still got visceral chills

[jcnporter]

Not sure if there's been an AI thread yet...opening a can of worms possibly..but I wouldn't personally use ChatGPT or anything similar.
Why not use this forum, feels like that's one of the reasons we're all here? - use this thread or start another one?
Post a band you want 'similar artist' recommendations for and I bet you get more interesting recommendations than any algorithm - might take a bit longer, though!

[Negamuse]

Yeah AI kind of isn't great for that and it's almost by design - any answer it could give you is derived from weighted averages of other people's opinions and by definition that's going to tend toward mass opinion.

The interesting stuff tends to be the stuff that's one person's favourite thing, that only they can tell you everything about

[Meddle]

Many hits of the past have been buried and forgotten. And many more non-hits. Go into your closest charity/thrift shop and sift through the 45s. Grab the ones you haven't heard of before if they are cheap. Or just look for them online (although a lot more than you think is not available online). You'll soon have a picture painted of what the most popular music of the past was, and the kind of thing people were making to try and score a hit - as opposed to what radio and pop culture have retroactively decided was happening. This is the most fun to do in the US where it seems like every little town had its own record label and local hits. Once you find some stuff you like, perhaps research it online and see who sampled/was inspired by it.
It might be time consuming but I'm sure you'll find it worth it if you love exploring music and own a record player.

[Alvin]

Current method:
- Promote into the places I know the least about.
- Gain new subscribers/followers.
- Consume their content.

I aim using paid promotions (I know that’s cheating) but there must be other ways to promote into the places we know the least about. Hit me up with ideas.

[Alvin]

(the idea being that a subscriber/follower can be the start of a mutual connection [I’m not good at this yet but I’m working on it]).

[Altruizine]

Yeah AI kind of isn't great for that and it's almost by design - any answer it could give you is derived from weighted averages of other people's opinions and by definition that's going to tend toward mass opinion.

That's not really accurate at all. You're only describing a certain primitive type of LLM that does what you described. A different form of AI that knows your music library and can, say, analyze and cross-reference audio files and detect specific patterns would be one of the best tools for this.

[Negamuse]

Yeah AI kind of isn't great for that and it's almost by design - any answer it could give you is derived from weighted averages of other people's opinions and by definition that's going to tend toward mass opinion.

That's not really accurate at all. You're only describing a certain primitive type of LLM that does what you described. A different form of AI that knows your music library and can, say, analyze and cross-reference audio files and detect specific patterns would be one of the best tools for this.

Hey if you've got a link to a system that can take music as an input and makes qualitative decisions about similarity that isn't based on

- IDing what you've given it as input based on matching fingerprints IE how Shazam etc do it
- comparing user preferences against known listening data like Spotify etc do it

I have a whole bunch of uses for that that's aren't limited to "find me more music like this"

I don't believe anybody has built that and I don't know how they would. The only systems that have actually ingested the world's recorded output and worked out how to do anything with them are things like Suno and Udio and they're not prepared to say "I made this, and I made it based on these songs" because of the legalities of it.