As Mr Carver demonstrated, by building an affordable amplifier, of which the sound could not be distinguished from an extremely expensive name amplifier that was being compared to (mid`s 70)!To clarify some points about that:
What Carver did, actually in the early and mid-80's, was to tweak the "transfer function" (the relation between output and input) of one of his amplifiers to match the transfer function of the highly regarded Mark Levinson ML-2 solid state amplifier (as chronicled in "The Audio Critic"), and subsequently to tweak the transfer function of another of his amplifiers to match that of a well regarded Conrad Johnson tube amp (as chronicled in "Stereophile"). His demonstration consisted, in addition to some ABX testing, of providing the two amps with the same input signal, and showing that when the output of one was electronically subtracted from the output of the other, essentially nothing remained. More precisely, a null of greater than 70 db was obtained, at least in the ML-2 comparison.
There were two major problems with all of that, however:
1)The tests showed, at best, that the two amps nulled against each other just with one specific speaker load, which was used in implementing his tweaks.
2)Credible anecdotal evidence subsequently emerged that he was not able to maintain anything remotely close to a 70 db null in production. Bob Carver essentially admitted this in an interview which appeared in "The Absolute Sound" about two or three years ago.
Regards,
-- Al