2

I downloaded a long DJ set as an mp3, and it contains a few hours of songs that I'd like to identify. The DJ won't be bothered to go through her set and send me a list, so I'm looking for a way to do this automatically. The songs are mostly in their original format; not pitch shifted or with additional beats

In the past I've been able to split longer recordings into individual tracks using some free software that detects silence in the recording. However, these songs are cross faded and so there isn't any silence between individual tracks.

Does anyone know of any online music identification services that can handle a file with multiple different songs?

brianerly
  • 21
  • 1
  • 2
  • I don't know of any, but the software in question would need to identify the songs individually first (like Shazam) and be able to repeat this feature multiple times and build a list. Some intelligence would be required as there would be no way (I think) of identifying cross-fades, other than that part not fitting any existing song. If you want to take the long route, use Shazam, then fast forward every time you've identified a song – nine9 Jul 08 '19 at 11:56
  • One solution I am considering is using R to automatically extract 20 second chunks every minute or so of the song, then submitting those shorter chunks to Shazam. I've got no experience working with Shazam's API (if it exists) or manipulating .mp3s in R, though, so I was hoping someone else had tackled something similar. – brianerly Jul 09 '19 at 20:14

2 Answers2

1

The Shazam app actually has an “Auto-Shazam” function (at least on iOS).

Hold down the Shazam “button” and it will keep listening and try to identify songs until you turn it off.

dijksterhuis
  • 139
  • 4
0

what song are in this mashup

https://on.soundcloud.com/k3JWE6cvxFkFUSXc9 will

user14493
  • 1