CHAPTER 17

Hashing Rock and Roll

17.1 OVERVIEW OF AUDIO FINGERPRINTING

You are a contestant on Triple Jeopardy and Alex Trebek asks “For $10,000, what performance group popularized this song?”

c17uf001

Is this Glenn Miller, Mozart, or U2?

Even though the reader might not actually be faced with such a question, a more practical application is provided by the requirements of The Music Performance Trust Fund (MPTF). Their web page explains that the MPTF is “… the world’s largest sponsor of live admission-free, professional musical programs.” Founded in 1948, the MPTF sponsors the public performance of free live music. An agreement between the recording industry and the American Federation of Musicians (AFM) sets forth the terms under which royalties on the sale and performance of recorded music are paid into the MPTF by recording companies. Audio fingerprinting is a technique to identify an audio signal; it makes it possible the following:

  • To generate efficiently playlists for automatic royalty collection
  • To verify automatically that a commercial for some useless, and perhaps even harmful, drug product has been broadcast according to the advertiser’s contact.
  • To enhance people metering; big brother/sister not only can watch you but also can check your musical tastes. Are you listening to Osama’s favorite jihad nasheed, “I’m Going to Kill, Kill, Kill …”?

17.2 THE BASICS OF FINGERPRINTING MUSIC ...

Get Hashing in Computer Science: Fifty Years of Slicing and Dicing now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.