I'm interested in doing some linguistics-inspired computational musicology on chord progressions and am wondering if there are any freely available databases of chord progressions.

Obviously, I could construct one manually from lead sheets, transcriptions with chords show, etc but am wondering if such things already exist.

By "database" I mean any structured data, preferably a text file or similar.

  • I am curious. What are you actually doing with this "linguistics-inspired computational musicology on chord progressions"? Are you writing software of some kind? May 19 '11 at 15:07
  • 2
    @musicwithoutpaper yeah, for analyzing chord progressions, inferring grammars and generating chord progressions; nothing too ground-breaking. Will hopefully help me develop more of my 'Sebastian' software too. May 19 '11 at 15:19
  • @James: Interesting. The reason I ask is because I am a programmer and I think it is possible and within reason to write a program to add a left hand to a melody that ands up sounding good... and I wondered how far that concept has been taken or if that was what you were talking about. That would be the part about "generating chord progressions". Glad this category has not been left untouched! May 19 '11 at 20:26
  • @musicwithoutpaper there's a decent amount of research out there on accompaniment generation (as well as melody generation) but, like many programmers, I want to develop my own particular take on it (and I'm a linguist by training so that gives another perspective as well) May 20 '11 at 1:56
  • 2
    Markov chains ftw guys. May 25 '11 at 16:39

The iReal b app for iOS uses downloadable chord progressions that are kept on the developer site's forums. iReal b uses a link that starts with irealbook:// and contains the chord data in the "URL" that follows.

View the source of the page linked above and you should be able to copy the whole thing.

  • 1
    The irealbook format requires cleaning up, its not insurmountable but you have to de-encode the string of every song - every space is a %20 etc and a bunch of other characters in it. Its a hassle, but doable. Aug 8 '11 at 1:58
  • 1
    In most programming environments, this is as simple as calling up a "URL Decode" function. This "%20" way of encoding characters for transport in URLs is actually a standard: en.wikipedia.org/wiki/Percent-encoding (this is more of a Stack Overflow question, really lol) Oct 3 '11 at 0:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.