I've tried Audessey many times and always thought it made things sound worse. One of the basic problems with receiver based Audessey systems is the lack of processing power necessary to run all the complex iterations. Essentially, receiver based Audessey programs takes a finite sample size and then plot results based on canned fuzzy logic algorythms. Those preset algorythyms usually come with the famous Audessey rolloff in the high frequencies. Something which I suppose may be beneficial in a lively reverberant room but which may screw up the speakers FR and introduce coherency problems and ringing. I can see there may be some benefit with non-linear speakers but not great speakers like Axioms.

Having said that I think the latest Audessey programs using notebooks to run the program look more promising. They take a much larger sampling size and can run the necessary complex iterations for much longer. I'm still not convinced that they can cure time domain problems. Indeed the real issue with electronic equalization is not so much in the frequency domain as in the phase shift artifacts they introduce to the signal.

A friend of mine in another audio manufactuer calls them speaker correction systems which have been effectively marketed as silver bullets. I've always gotten better results by leaving Audessey off and playing with room treatments, seating and speaker positions.

As always YMMV and if it works go for it. Or as Alan usually zings when we can't resolve a difference of opinion, " If that's the reality you want to believe, that's OK too."

John