Don't know if any of you people know anything about EchoNest. Apparently you can upload an audio file, and you will receive back an XML file containing data about the audio... beat detection, 'melodic content' and dynamic range.

Is it possible to read an XML file in PD, and then use the info contained?

I've been looking at this too.
Have you had a look at one of the output XML files yet? The beat detection comes as a list of "Tatums" in seconds. I would have thought you could easily read this to an array once you've got over the format problem right? Exciting!?

check here for more on the information provided by the analysis:

http://developer.echonest.com/docs/xmldescription