My understanding is that the graphs were measured in the NRC anechoic chamber, which is accurate down to 80 Hz but rolls off below that. Apparently this is a function of the size of the room; bigger chambers are accurate to lower frequencies. Axiom combines the NRC anechoic chamber measurements with outdoor "speaker up on a pole" tests to get the overall frequency response.

My understanding is that the dip starting around 80 hz is largely an artifact of the chamber size but I'm not 100% sure about that.

EDIT - here are some comments from Ian about this :

http://www.axiomaudio.com/boards/showflat.php?Cat=&Board=tech&Number=89253&page=&view=&sb=&o=&vc=1

Last edited by bridgman; 09/24/05 03:18 PM.
M60ti, VP180, QS8, M2ti, EP500, PC-Plus 20-39
M5HP, M40ti, Sierra-1
LFR1100 active, ADA1500-4 and -8