Here is some early Dr. Xin insight into what gives with the Reference:
"I did think about the plug-in of the huge cap, the problem is: LE is not just an added cap, many components need to be changed correspondingly, to ensure the integraty of the power suppply, charging and battery maintenance circuit.
LE's key is the lower voltage that allows for the use of AD8656, AD8616 5.5V OPAMPs. More and more people love the sound of these chips with SuperMicro-IV and SuperMini-IV, they also love the 4 switches, metal case and fast charging of SuperMacro-IV.
After removing and replacing the diodes in SuperMacro-IV with the new ultra-low impedance FETs, the power impedance is already all way down to 0.1 ohm (from 8 ohm!). So if you use AC/DC, the impedance is already down to the bottom. A 12V regulator's impedance is only 0.024 ohm, making the total still less than 0.2 ohm.
The bottleneck is now the batteries. NiMH's internal resistance is 0.2 ohm (0.17 after fully charged and 0.18 after discharge, to be exact). 8 AAA will be 1.6 ohm, much higher than 0.1 ohm, but still very low to a 32 ohm load. Reducing to 4 AAA, it is only 0.8 ohm. As the room is there, so why not add a 15000uF, to drag the impedance further down to 0.3 ohm at 20Hz. Such kind of power impedance is insanely low, even for a home speaker amp. Is hi-end audio all about over-kill?" from:
http://www.fixup.net/talk/topic.asp?...hTerms=15000uF
Here is another recent post concerning the effect of amp impedance vs. headphone impedance and its desirability.
http://www.head-fi.org/forums/showpo...&postcount=299
I find it interesting what Dr. Xin is doing with his Reference Amp. I may have to spring for one.