Live data from Hacker News

The fake artists problem on streaming platforms

tedgioia.substack.com

51–60 of 160 posts

Re: The fake artists problem on streaming platforms

#51
I call this the Pentatonix Problem at Christmas. When left in "auto-play" mode, streaming services absolutely default to the lowest payout possible. The fastest way to discover if there was a Pentatonix version of a song is to ask "Hey Google, play " ... "OK, playing by Pentatonix"

So yeah, they pay a bunch of basement recording artists to throw together a million generic sounding low-effort Ableton jams and voila, lower play counts and lower royalties for the normal working artists.

Some genres are easier to game than others. Classical music is really rough right now. Unless you already know EXACTLY what version of the piano concerto you want, you're going to get some crappy midi playthrough that was likely just the "performer" dropping the sheet music straight into the midi roll.

Re: The fake artists problem on streaming platforms

#52
post #47

What makes them "fake"? I kept waiting for the author to explain how people are taking others' music and re-packaging it as their own to collect royalties, but it doesn't sound like that's what's happening. SEO-optimized music is not "fake music". It's an interesting phenomenon, and maybe even a bad one, but framing it this way is disingenuous and alarmist.

What makes the artists fake is that they are publishing only a couple songs for the specific purpose of providing the streaming platform with royalty free music, and then they disappear. The example pictures in the article show artists that have published 6 minutes and 8 minutes of music. There's also a statistic that 20 people created 500 artists. These artists are shells, facades, illusory, fake. Go looking for the…

> What makes the artists fake is that they are publishing only a couple songs for the specific purpose of providing the streaming platform with royalty free music

The article has 3 examples. Two have easily identified 3rd-party labels, the last is unclear. Neither is a traditional label, but there is also no indication they are owned by Spotify.

> People buy it expecting a curated playlist of music they want to hear, and instead what they get is a curated playlist of music that sounds similar to what they want to hear.

What bands would people expect in a playlist of background music? In fact, one is from a provider of royalty-free background music - which seems like exactly what you would expect in such a playlist. You could argue that it's not really jazz, which is understandable, but I'll file that gripe along with my anger that "New Metal Tracks" features lots of metalcore.

Re: The fake artists problem on streaming platforms

#53

If generic jazz can easily substitute "important" jazz it just means your important jazz doesn't have the value you ascribe to it, especially not in the context of "things for background music." I don't think this is saying anything about more active music consumption where people are actually listening to it. I personally get annoyed if more than 20% of a mix is songs I haven't already upvoted so there isn't a lot o…

Music historians note that the creation of the "cannon" of classical music occurred when public orchestral concerts arose around the early 19th century (and often paired with the rise of the middle class). Prior, you'd need an invitation to a private concert put on by a noble. Mostly "old" music was played at these concerts. A public concert had to at least cover its cost from ticket sales, so eliminating commissions…

Canon? (Unless you're talking specifically about the 1812 overture. :-) )

I wonder how copyright extension has affected this phenomenon. Works taking decades longer to enter the public domain, leading to the existing public domain (old) music becoming even more solidified as classical canon? If anybody knows about this I'd love to hear more.

Re: The fake artists problem on streaming platforms

#54

What makes them "fake"? I kept waiting for the author to explain how people are taking others' music and re-packaging it as their own to collect royalties, but it doesn't sound like that's what's happening. SEO-optimized music is not "fake music". It's an interesting phenomenon, and maybe even a bad one, but framing it this way is disingenuous and alarmist.

My assumption here is that popular songs are covered then assembled into generic playlists which are then used to divert the lion’s share of revenue away from the original artist’s music. Thus providing Spotify or the labels a low-cost way of minimising their royalty payout figures for listeners that aren’t fussed if they hear the original or a cover (eg mindless background music).

They are “fake” artists as they’re hired by Spotify, the label or a 3rd party to produce the cover for a fixed fee, which the rights are then held. To use various names also distracts from what would be highly repetitive production credits.

The author notes that such investigation isn’t possible on services other than Spotify because they do not provide much background information such as play counts. However this information is not necessary, one can simply play the generic playlists made by the music services and check to see if they are the originals or covers.

There is definitely large variation in the streaming industry here. It also wouldn’t surprise me if music services are favouring these playlists as a means to reducing their royalty payout costs.

Re: The fake artists problem on streaming platforms

#55
post #31

If generic jazz can easily substitute "important" jazz it just means your important jazz doesn't have the value you ascribe to it, especially not in the context of "things for background music." I don't think this is saying anything about more active music consumption where people are actually listening to it. I personally get annoyed if more than 20% of a mix is songs I haven't already upvoted so there isn't a lot o…

If generic jazz can easily substitute "important" jazz it just means your important jazz doesn't have the value you ascribe to it, especially not in the context of "things for background music." This is a modern version of Seeburg background music. Seeburg was a jukebox company, and as a sideline, they also sold a background music system. This used a special purpose record changer that played a stack of records over…

I note that streema.com is trying really hard to get me to install "easy search tool" by favouring the START button above the PLAY button. If it's not them directly, it's a dark-pattern ad they are running

Re: The fake artists problem on streaming platforms

#56
A few months ago my toddler was in a "must have music" phase, and when we got tired of switching records or listening to the same local radio ads over and over, we turned to Spotify. The kids' music playlist we stumbled on had one album with bizarre choices of tempo, instrumentation, and effects. It kind of grew on me after a while.

I wanted to learn more about the artist but when I tried searching (it was a super generic name like "Music for Kids") they only seemed to exist on Spotify. I'm disappointed but I can't say I'm surprised.

Re: The fake artists problem on streaming platforms

#57
post #24

I'm not sure I understand the author's point here. Spotify is recommending artists outside of the top 40 charts who they haven't heard of before, therefore they are fake?

There are steps missing, perhaps -

Artists with only 2 songs on Spotify are probably fake, there are odds they are the property of Spotify themselves and have been carefully created and placed into playlists to minimize royalty costs to Spotify.

I love to discover a great new song and artist on a playlist, where that artist has a whole bunch of songs for me to try out - such an artist is much more real. And I'm willing to pay a tiny bit more to access real artists because of that.

Re: The fake artists problem on streaming platforms

#59

It’s shameful to read all the folks claiming that actually known famous historical jazz greats are somehow being legitimately usurped by generic nothing music even in the context of background music. What the article discusses is clearly dishonest, whatever your take on the word “fake” Grammy awards are indicative of some popularity and reknown, so not seeing Grammy winners on Spotify genre pages is suspect. There ar…

I don't see anything remotely dishonest about this. "Spotify genre pages don’t reflect these but push generic in-house brands as top tier products on the world stage" is AFAICT not based on reality. Spotify's playlists for the hottest hits in the genre or decade of your choice have exactly what you would expect them to have. The generic background music only shows up when you ask for generic background music.

Re: The fake artists problem on streaming platforms

#60
post #52
post #47

Earlier quoted context omitted.

What makes the artists fake is that they are publishing only a couple songs for the specific purpose of providing the streaming platform with royalty free music, and then they disappear. The example pictures in the article show artists that have published 6 minutes and 8 minutes of music. There's also a statistic that 20 people created 500 artists. These artists are shells, facades, illusory, fake. Go looking for the…

> What makes the artists fake is that they are publishing only a couple songs for the specific purpose of providing the streaming platform with royalty free music The article has 3 examples. Two have easily identified 3rd-party labels, the last is unclear. Neither is a traditional label, but there is also no indication they are owned by Spotify. > People buy it expecting a curated playlist of music they want to hear,…

I definitely had the wrong impression of how much music was produced by the examples given. It's not just "two songs and dissolve".
Post reply on HN