Speech signals modulation entropy

Hi, 
I got a question regarding to modulation entropy of audio signal. There is
some literature I found about this topic,
http://www.irit.fr/recherches/SAMOVA/FeaturesExtraction.html#me.

I've already read sample .wav file. But now, how do I calculate entropy of
this sampled signal with the lenght of N samples? How should I understand,
in this case (in paper equation from above), propability of event?



---------------------------------------
Posted through http://www.DSPRelated.com