Seems cool and I understand you're getting hugged to death by HN right now but I could not use it because the search was too slow and as soon as I change focus to another window for example, it removes whatever I wrote in the search bar so unfortunately completely useless due to this bug.
Show HN: I trained an AI model on 120M+ songs from iTunes
231–240 of 444 posts
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#232I find it disheartening that the comments are loaded with people saying how bad it is. The interface is awesome, the ability to have search results pop up and offer suggestions is impressive. I thought the results for a Lana Del Ray song and a song by Carlos Santana were both interesting and suggested new things. We should do a better job of supporting creators here.
Thanks! Definitely welcome all the feedback here.
Gonna ship a better, improved model soon!
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#233Pretty cool. I think I broke it with Enya's Pax Deorum , though... can't get it to work at all now ("Hmm, something went wrong.")
Sorry! The service is sometimes bugging out due to the recent surge in traffic, which is causing some songs to consistently fail.
Things should become much more reliable and better soon!
Also, sometimes there's a delay between when a song is available in the text search bar, and when it has been indexed with my AI model (which can also cause this error).
This will also be fixed!
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#234I'm wondering if anyone has done something similar, but instead of trying to find similarities in the raw audio, they use tags available from sources like Last.FM, Musicbrainz, Discogs etc? And the ultimate answer to that is probably "those sources kinda suck". Discogs is like a trainspotter on the spectrum, fascinated by release IDs. Musicbrainz is kinda similar (each song will have a dozen matches of wildly differe…
Discogs is great, it just doesn't concern it self with how the music sounds...
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#235A good chance OP is going to hear from Apple legal, Apple recruiting, or both =) But cool project!
:)
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#236Interesting, I put in a bunch of my favourite songs and found nothing I liked. The songs seem to be over matching on drum beat and tempo so there is a similarity to the song I suggested and the matches but it's often superficial. For instance I picked a song with a very strong snare drum line. All the suggestions also had a strong snare drum line but wildly different melodies, genre's, tone, etc.
Yea, this is due to a shortcoming in the current model's design.
Got some stuff in the works for an improved model, hopefully will be able to ship it soon!
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#237Interesting. Some feedback 1. Please add match score 2. Group and fold duplicates 3. Add the year with the sort feature - to identify rip offs
Noted! got a lot of updates shipping soon.
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#238I entered "Spiro - The Vapourer", a brilliant dance music inspired instrumental folk piece with intense, layered melodies. It recommended "Buzz Cazon - Sentimental Attitude". I have trouble putting into words how bad that recommendation was. Seriously, just compare those two pieces, do they have ANYTHING in common besides maybe vaguely the tempo?
OK, but I'll concede it does SOMETHING pretty wild. Just not necessarily something good for recommendations.
Because I tried another song: "Sam Sweeney - The King of Prussia's March", and the second recommendation was "Polsk Nr. 48" by Rasmus Storm.
The wild thing? Well, first of all, Rasmus Storm isn't strictly speaking the artist. He was the composer. He was a 17th century Danish fiddler, who - rarely for his time and social standing - knew music notation, and wrote down his tunes in a book.
The crazy thing is that in that same book, "Murchy nr. 14" is the tune better known in England as "The King of Prussia's March"!
The odds of that happening by chance has to be absolutely tiny. "Obscure" doesn't even begin to describe it when it comes to Rasmus Storm's notebook.
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#239I entered "Spiro - The Vapourer", a brilliant dance music inspired instrumental folk piece with intense, layered melodies. It recommended "Buzz Cazon - Sentimental Attitude". I have trouble putting into words how bad that recommendation was. Seriously, just compare those two pieces, do they have ANYTHING in common besides maybe vaguely the tempo?
Wait, you really don't hear the similarity? To me the two pieces sound uncanny close. The tempo sure, but also the rhythm guitar does almost the same thing, and the higher pitched instruments in both have similar timbre and do similar things. You could probably mix the two pieces together and it'd sound alright.
This happened to me before, I'd point out that for example Reckoning Song and Counting Stars are the same song, but people would swear they have nothing in common. Or Sail vs Believer, why was the second one even made. Is it that people focus mostly on the manner/vibe of a song, and don't notice what it actually sounds like?
Re: Show HN: I trained an AI model on 120M+ songs from iTunes
#240Seems cool and I understand you're getting hugged to death by HN right now but I could not use it because the search was too slow and as soon as I change focus to another window for example, it removes whatever I wrote in the search bar so unfortunately completely useless due to this bug.
my bad, that can be annoying. will ship an update for this!