Audio is very complex but so far we have been able to corelate the listening experience with the science. There is no perfect amp yet and everyone has a preference. So, even for a very good amp such as NuForce, we find that a lot of people raved about it because of a good bass, sound stage etc. High freq imaging WAS a weak point (compare to some of the best) and it has been pointed out by a few people. Sure, there are many who find that it is good enough due to their setup, music preference, ear characteristic or whatever. We enjoyed bathing in their praises

. But the science doesn't lie. I wonder why icepower datasheet doesn't show THD+N vs Freq at various power rating

. And the spec only showed THD+N at 1Khz, etc up to about 7khz. hmm...
For those who follow the plots, each upgrades showed significant improvement in measurement and that corelate to listening feedback.
With the latest upgrade, we have taken measurement that showed THD+N at 0.03% at 20khz at 10W for example, with a 40khz filter (ie no cheating). As a comparison, the version that was first audition by some people has THD+N at 0.16% at 20khz at 10W (even this number is highly respectable). I doubt any other switching amp or few linear amp can achieve 0.02x% THD for 20Khz. Recent listening test have confirmed the improvement in high freq imaging.
But good sound reproduction is more than just THD+N and freq. Harmonic and power plays an important part too. We didn't just bring down the distortion and noise to an astonishingly low level (we are still trying to confirm our SNR of 90+ to 100db and that's amazing for a switching amp with switching power supply), harmonic has cleaned up. Many people wondered why NuForce with 100W rms sounded as loud as some 300W linear amp and that's due to the fast response time of the SMPS.
Someone demo NuForce Ref 8 recently at a show in Chicago (related to AV123) and someone was wondering where did we hide the tube.
NuForce has zero phase shift and near zero output impedance at ANY frequencies and that's the reason it matches well with many speakers. We know that Current hungry speakers (Apogee) will not be a good match for Ref 8 due to its SMPS Current limit of 2.1A. The amp's power can be estimated with V square / 2*load but SMPS will current limit even though it seems that there is enough power if you rely on voltage for power calculation. To address this problem, we put in a 350W RMS SMPS in Ref 9 with voltage limit at 52V and current limit at 7.3A.
With all the rapid improvement, we're exhausted and need to chill out for a while
