0

I'm creating an app that recognizes the songs of some web radio feeds, I would need to parse feeds to find songs, I found a paid service (ACRCloud) that allows you to do so, but the cost is high given the number of radios that I would like to analyze, so I searched a bit and found Echoprint, I could use it for my purpose? Would be suitable? Why don't I find the documentation ... I don't know, maybe there are better solutions? Thank you

user6634742
  • 1
  • 2
  • Questions asking us to recommend or find a book, tool, software library, tutorial or other off-site resource are off-topic for Stack Overflow as they tend to attract opinionated answers and spam. Instead, describe the problem and what has been done so far to solve it. – André Kool Jul 25 '16 at 11:28

2 Answers2

3

Echoprint would generate good fingerprints for your usage (exact copy of musics, recognition time in the scale of tens of seconds).

However, you will have to maintain a database of known musics. Which is difficult to keep up to date. The mothership of Echoprint, the Echo Nest, offered music recognition in the past, but this seems now over. Though, a server designed as fingerprint storage and retrieval is free to use.

What you could do to generate a database is:

  • fetch metadata from some streams, when available.
  • download the youtube version of those musics. See the youtube-dl tool for that (disclaimer, this may be violating the TOS of youtube)
  • inject fingerprints in the echo nest server.

This could bootstrap efficiently your system. But you may have a hard time reaching the performance of paying services like ACRCloud and similar. It depends on your requirements.

astooooooo
  • 323
  • 2
  • 11
0

We run a service for our website which follows 9 radio stations, using the internet stream of those stations. We show real time what is playing on those stations.

The library (in c#) and a database of 1.3 million fingerprints can be found at:

https://github.com/nelemans1971/AudioFingerprinting

Example programs include the software to follow a radio station.

leo337
  • 46
  • 2