Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

SomaFM

somafm.com

Post reply on HN
Print view View original

161–170 of 170 posts

Re: SomaFM

#161

Post by Anthony-G »

rbanffy wrote: post #132
huron wrote: post #64

Earlier quoted context omitted.

But how do you find new songs for you to put on your playlists? That’s why SomaFM and online stations like KCRW are so valuable. They’re ways to introduce new music and even new styles of music to the listeners.

Here in Dublin we have Raidió Na Life, which has the best music programming of any "offline" radio. Everything spoken is in Irish, of which I understand almost nothing, but the music, is amazing. Online, of course, SomaFM has the best streams.

Upvoted for reminding me how good Raidió na Life used to be. I haven’t listened to it in over 10 years. I used to know the Neuromantek DJs back in the 2000s. I’m amazed to see that their show is still running! These days, when I listen to the radio, it’s usually via Saorview but I must make the effort to hook up Raidió na Life and other radio stations to my home hi-fi (now that my Squeezebox has died).

Re: SomaFM

#162

Post by meanmrmustard92 »

rektide wrote: post #5

I really miss Shoutcast/Icecast being such an active part of my life. The quality is way down, and there's like 1/6th the channels there used to be. For electronica in particular, there was just so much. But there were also so many random niche stations! SomaFM continues to be a great place for highly genre-d music. I make sure to give em money semi-regularly. Suburbs of Goa[1] is one of my favorite. [1] https://soma…

Shpongle and OTT are amazing goa-ish electronica. OTT's new album is great.

Re: SomaFM

#164

Post by jrnichols »

I still have bookmarks for groove salad and occasionally use them - on a 1.25ghz iMac G4 from years ago. The old Mac OS 9 version of iTunes still works and still plays internet radio streams.

Re: SomaFM

#165

Post by godDLL »

johnchristopher wrote: post #125
habi wrote: post #116

I have these three aliases in my ~/.bashrc: alias somafm='mplayer -really-quiet -vo none -volume 128 -playlist http://somafm.com/groovesalad.pls' alias goa='mplayer -really-quiet -vo none -volume 128 -playlist http://somafm.com/suburbsofgoa.pls' alias beatblender='mplayer -really-quiet -vo none -volume 128 -playlist https://somafm.com/beatblender.pls' Thanks for the reminder to donate again!

Nice :) Doesn't `volume 128` saturate sound though ?

I have a 2009 Mini connected to a Creative Labs 2.1 and it does.

But then, there's a Lenovo Yoga from like 2016 that plugs into a pair of SOLO7c and that doesn't.

Different stations I start with volume 55-85, they don't all have the same loudness (genre related, no doubt). But the use for radio is for me "fill space, drown out distractions, provide flow" so that I don't need it loud. I'm not actively listening to radio. I didn't choose the song. And that's the point.

Re: SomaFM

#166

Post by johnchristopher »

godDLL wrote: post #165
johnchristopher wrote: post #125

Earlier quoted context omitted.

Nice :) Doesn't `volume 128` saturate sound though ?

I have a 2009 Mini connected to a Creative Labs 2.1 and it does. But then, there's a Lenovo Yoga from like 2016 that plugs into a pair of SOLO7c and that doesn't. Different stations I start with volume 55-85, they don't all have the same loudness (genre related, no doubt). But the use for radio is for me "fill space, drown out distractions, provide flow" so that I don't need it loud. I'm not actively listening to rad…

> I'm not actively listening to radio. I didn't choose the song. And that's the point.

Who are you arguing with ?

Re: SomaFM

#167

Post by kleiba »

rbanffy wrote: post #131
kleiba wrote: post #66

Earlier quoted context omitted.

Isn't that what recommender systems are for? /scnr

They can only recommend what they understand. I am not aware of any automated recommender that listens to music to make decisions on how to recommend it. For that, humans are still good.

They don't have to listen to music - they just look at what people who like similar stuff to you, like in addition to that.

Re: SomaFM

#168

Post by aparticulate »

tbatchelli wrote: post #32

SomeFM is such a rarity these days, a reminder of how we believed the Internet was going to be. They take recurring donations. Probably the best way to help them and one of the best investments for me, as keeping it up and running is a way to keep these naive ideas alive. The music is curated and is most excellent (for my taste, of course). I have Groove Salad on speed dial in my HomePods around the house (through Tu…

> a reminder of how we believed the Internet was going to be

very well said

Re: SomaFM

#169

Post by gffrd »

inferiorhuman wrote: post #115
hfourm wrote: post #80

Earlier quoted context omitted.

To be honest, Pandora has served that role for me over the years. I am always amazed to look back at my Pandora station history to see how it has evolved into different streams/genres, all stimulated from hearing new music through a Pandora channel, and then starting a new station after I liked it. This has created a web of new music I wouldn't have sought out otherwise. Obviously, I do think that a human DJ may perf…

My favorite part of SomaFM is that I discover new artists (or get encouraged to go on a deep dive) fairly regularly. I've never bothered with Pandora or Spotify but the impression I get is that SomaFM is much better with the more niche genres.

I find the same. I feel more "weight" in a choice made by a human, who has their own "map" of the music and what connects it … so, I care to know about the choices made and why they were — the context of the choice.

I'm happier with a song pairing that has some arbitrary, personal history than one that's a logical walk down a relationship model.

Re: SomaFM

#170

Post by rbanffy »

kleiba wrote: post #167
rbanffy wrote: post #131

Earlier quoted context omitted.

They can only recommend what they understand. I am not aware of any automated recommender that listens to music to make decisions on how to recommend it. For that, humans are still good.

They don't have to listen to music - they just look at what people who like similar stuff to you, like in addition to that.

And that's why a human (or a model that can represent the music) can make better recommendations.
Post reply on HN