0

I am developing an application which listens live streams and fingerprints them concurrently in order to be identified by a mobile phone.

I will use gracenote's sdk for fingerprinting process. Since I need to fingerprint live streams instantaneously and get rid of them after a few minute,I need to design my own server and database (I also do not know whether you have such server in your sdk or not.).

Now I would like to ask a question about your fingerprint, I can get fingerprints successfully with gnsdk_musicid_query_get_fp_data function in base64 format.

First question: What is the length of coded strings, which are base64, in seconds.

Second question: After decode base64 codes, How should I do the comparison, in binary wise(after concatenation of bytes) or in integer wise, if integer wise how many bits to align (1,2,4 or 8 bytes?) .

Could you give some hints about comparison?

mustafa.yavuz
  • 1,274
  • 2
  • 21
  • 40

1 Answers1

2

The GracenoteSDK only has the ability to create fingerprints. The only implementation that can compare fingerprints is the Gracenote Service. So you won't be able to make a service of your own that will do fingerprint recognition, but you can certainly build an application that uses Gracenote to do this.

PQuinn
  • 992
  • 6
  • 11
  • In this [post](https://developer.gracenote.com/audio-recognition-server) it seems we are allowed to do this. I need special purpose server (fingerprinting live streams) so I had to implement my own. I just need format of the fingerprint. I do not understand why I won't be able to make my own service. Is it about license conditions? – mustafa.yavuz Sep 04 '13 at 06:21
  • It's not about licensing and such. It's just that right now there is no way for you to build a local database filled with your own audio samples. It's simply because the code pieces you need are not part of the SDK. Only the Gracenote service is able to compare the fingerprints to our reference fingerprints database. As you can see in the link you post, Paul is saying the same then I do in his second answer. – Jonx Dec 26 '13 at 22:12
  • Perhaps you need to look into some alternatives like EchoNest or AcoustID. – liftarn Apr 23 '14 at 11:59