VAC - any comments? Good or bad.

Part II (Sorry!)
I guess another part of this is that manufacturers generally use resistive loads for amplifier measurement, wherein linearity usually appears reasonably flat. We're now used to seeing a curve now and again from some testers (like Atkinson at Stereophile) who throw in a "projected behaviour with an 8-ohm real speaker" curve, which, especially with a high output impedence amplifier, will show some non-linear frequency dependent behaviour.
In a way, there's some useful "single-blindness" in having very raw data. Indeed, after running the first SS1 vs VAC test, one can't tell by looking at the raw data anything.
I then ran the second test: SS2 vs VAC. A cursory examination of the two sets showed that SS1 very nearly equalled SS2, and that VAC very nearly equalled VAC. I then tried SS3 by itself, which equalled the two SS. The amplitude of difference between the VAC and the three SS was quite large (see the curves again), and correlated nicely with the gross differences in sound differences. Certainly the three SS amps had differences in sound, but these had to do with grittiness, edginess, decay, dimensionality, and all that typical SS stuff. The Audio refinement Complete Integrated clearly sounded the best. The Acurus and NAD I didn't like much. But ALL THREE SS AMPS sounded TIMBRALLY EQUAL, and indeed measured spectrally the same. It all makes sense. (Now do you understand, Balekan?) The three SS curves are very tight (intertest imprecision mainly given only by pink noise bias difference set by gain control). The two VAC curves are quite tight, only being a bit at variance at the 16k and 20k points where both are at highest slope, and therefore coarsest measurement...but also greatest non-linearity--quite wild rides....
The only almost-sily fly in the soup here is that someone could actually postulate that since the raw data is relative to an unknown (yet stable) reference base, isn't it possible that it's the VAC that is linear, and that the THREE SS AMPS are all wrong? Yup. Entirely possible. And if there was only one SS amp I'd have to agree, and would never have published these results. But when THREE different SS amps agree so highly, and one TUBE amp disagrees so greatly, TWICE, and sounds horrible, to boot, my money's on something terribly wrong with that output stage...when mated with the supposedly benign Fidelios. Ern

Ern,
"...and please not be coy about summoning the industry gods to pass judgement on what anyone of reasonable facility with the scientific method can discern is a well-run set of differential tests..."
Why would you refer to industry "experts", as "Gods" in such a sarcastic manner? What ax are you grinding, and why?
My comments were aimed at the potential of some possible unusual load of the the speaker, and possibly room effects to skew any results, and that if you used a host of speakers, that the test results would be different with each differing reactive, rather than resistive load.
There are numerous questions, but a few come to mind. Did you use calibrated mic's? Have you previously plotted the room for additive frequency anomolies, or did you do as most real qualified scientists do, and use an anechoic chamber, or a faux chamber, designed with the flaws input into the response curve of the room,in order to factor those errors out? Again, I would ask if you use more than one set of speakers, if so were the results of the amps exactly the same?
No one is peeing here, I simply said I thought you did the tests with sincerity, but that I doubted the validity because the reaction of only one set of speakers, with one Avatar amp, and no pre test,test data on it, to compare with others of its ilk,(meaning, was the amp tested, and confirmed to be in perfect, or its intended operating condition and certified to be so?).
Speakers present reactive not resistive loads, as everyone knows, hence, tests of an amplifiers output, is done with various resistors, (still flawed) yes but equally so, since this test is still into resistors, even though of varying values.
Cable capacitance, inductance, any of that factored in, or would you say that this doesn't make any difference whatsoever in your tests?
Why are you buying such flawed gear? If you have doubts about a product and its output, why not try before you buy; why lie in the bushes (metaphorically) and ambush any manufacturer with such (imperfect, incomplete) testing. Even though I have written many press releases that were released nationally, that does not make me Ernest Hemingway.
And just so you'll know, even though the folks at VAC didn't really call you to task on this test, I know personally, that a coffee mug was offered as a prize to anyone who could explain your tests, and how they relate to real world.
The implication of your test is that the Avatar is flawed in its output, and that Solid State amps are not. I don't believe that, as a narrow finding, that this is correct. I don't know, and neither does anyone reading this, that the Avatar in question was working correctly when you "tested" it.
Also, a little presumptious to think you can educate someone you don't know, and even offensive based on this writing you give as support for that claim.
If only Kevin Hayes weren't so concerned about public image he would join this discussion and "enlighten" you, as to the variables which are not dealt with or even mentioned in your diatribe.
Again, I am only one person. Ask any (and they aren't God's that is insulting) industry expert how valid these tests are, don't ask or try to convince me.
By the way, what audio products have you designed? How well do they sell? With your obvious tallent you should teach all these manufacturers the error of their ways. You seem to be a disgruntled, anti audio person, on some not so obvious quest to discredit VAC. If that is the case, I suggest you move on, sell it and forget it. If not, please be more respectful.

Have owned Renaissance 140 mono blocs from VAC.Still sounding great and reliable after 10 years of hard work driving my WATT speakers.WILL NEVER SELL THESE AMPS!VAC stuffs are expensive but WILL SATISFY YOU. Khai.

Oh ya my sentiments exactly.....I have had many pieces from VAC over the last 15 years, the most recent and current is the 70/70 III amplifier. There is no question that VAC absolutely makes some of the most musical gear available for the dollar-regardless of price. And their design and craftmanship are second to none IMO. And just as important to me when buying expensive high-end gear is the commitment of the manufacture to support his/her product and I can tell you from my own experience that the customer service at VAC has always been nothing less than stellar! Kevin and his team has always, and continues to, make every effort and sacrifice to support his product and make his customers happy. In my opinion VAC exemplifies all that is great in this fun hobby.

Lrsky, you still don't get it! I have absolutely no axe to grind, DON'T OWN the bloody amp, and will try to answer your queries in order, as I flip back and forth: I used the terms "gods" because the suggestion that CEO's be called in to assess a simple differential test is technically naive. Any of the mentioned companies techs well-grounded in the scientific method would pass on the test without mention. Seriously. You simply do not see this. Secondly, the Verity Audio fidelio poses no unusual load whatsoever, being of fairly flat impedence with a 5.7ohm min, I think. But OF COURSE it's expected that the results would be different with a change in speaker IF the output impedence of the test amp was high. That's precusely the test's hypothesis! Why the bugaboo that the results simply prove convincingly that to be true? Again, changes in LOAD, and room loading, areinconsequential, as they are FIXED across all trials. Again, the basis for a differential series of test runs: only one variable (the amp) changes. Please try hard to understand the reasoning behind this. No condescension meant. It's critically important....Only one mic was used, so it's calibration is not of concern. Do you understand why? If not, I can explain later. Again, as the room anomolies remain FIXED, their effects are NOT of concern. Your dig was not necessary. I've TAUGHT more real scientists than you can count. No, I did NOT use more than one set of speakers (one per channel), because the point was NOT to investigate HOW the VAC varies with speaker load changes, which is another experiment requiring different test equipment. My test procedure was only to determine what the DIFFERENCE in frequency response was between the VAC AVATAR and the two SS amps I had on hand, as empirical audible evidence indicated a great discrepancy....No, I did NOT in any other way test the VAC beforehand, as it was a dealer demo, and naively trusted to perform normally. Both output halves matched, which seemed a rudimentary qualification, as well. It's not a customer's job to perform manufacturer's final QA. Yes, I'd rule out the 8' Nordost Red Dawn cables' very low inductance and resistance as significant contributants here. But that suggestion simply muddies the waters again! All system componentry (test equipment and conditions) were listed, unvaried, and NOT unusual in any way such as to skew freq resp between ss and valve amps, as most will agree....Again, I didn't "buy such flawed equipment", and thus didn't "lie in the bushes to ambush" anybody. I simply doubted what a gushing salesman who sold me Verity Audios told me about the Avatar, as my ears heard differently, so I tested it against two other amps in a very simle, but VERY VALID manner, and produced extremely significant results. Certainly NO-ONE has countered the methodology to date. A full list of the raw data is available to VAC as well anyone, although I doubt that anyone cares at this point, 4 yrs and a redesign of the amp later!...Again, why is my testing incomplete and imperfect, Lrsky? I remember back in the mid 80s I sat as the US rep on a Technical Advisory Group at ISO Tech Com 42 (Lab Equip) in Frankfort, where the DIN Secretariat lauded us for writing a standard that was efficient and tight, containing only what was "sufficient, but necessary". It was important him, as he had to organize translation into 17 languages for balloting. If I had tried to reinvent the wheel with every procedure we proposed in an ISO method we would never have gotten anything done! So sir, you may or may not be another Hemingway, but when it comes to running a simple differential test, please just trust that I know what the hell I'm doing. And I'm quite sure that a coffee mug can look like a red herring from certain clever angles, and necessarily so, as it's easier to shoot the messenger and leave the waters muddy, as you've done....
However, it IS true that NO ONE knows if the suspect AVATAR was working correctly when tested. I do remember that it sounded the same in BOTH channels, in both triode and pentode mode. So I continued like an ordinary consumer, trusting that it was a normally-functioning unit. It had ample output, too....
Lrsky, your last condescending remarks are is absurd, and perhaps psychological projection? I need ask no one if my testing methodology is valid. Chrissake, just take any college-level lab course and learn how to perform a one-variable test procedure. It's not alchemy! I've designed a couple of two-ways, and a three-way in the early 90s, using the same test setups to carefully tweak crossovers in 1/3 dB increments over 1 octave bands, thereafter learning you can't get a manufacturer to supply you with driver pairs matched well enough, and I didn'rt want to get into computer-matched pairs. I have GREAT respect for Verity, Snell, even Boston, et al, who control driver spec closely enough to make nice clones. I learned this from many years controlling production and QA of the world's most popular laboratory volumetric measurement equipment, called the Gilson "Pipetman". It's a handheld small volume auto-pipettor used down to 1 uL, calibrated by differential test methodology I developped in the 70s using gravimetry at the
1ug sens level. It's this stuff that I honed into ASTM and ISO standard procedures for calibrating volumetric ware so users in labs worldwide could test their pipettors, dilutors and dispensors inhouse efficiently instead of having to buy new ones, or, as you would prefer, Larsky, measure the room, all the beakers, and maybe the drapes too, and perhaps ask for permission before seeing if it's ok! Sigh...why the doubting Thomas?
Of late my semi-retirement and slipped disc has me still enjoying life in a third career as Boston's SubaruGuru, and as a contributor to this community, supplier of nifty $39 DIY all-Teflon PCKits that keep me in touch with a couple hundred of you guys....
Disgruntled? Hardly. Anti-audio? I've 50-60 cheap CD's I scooped up in Italy and at Heathrow this August I haven't even goten to yet. My two-ch ref system and my Steinway B are the aural delights of my life (plus frequent visits to Jordan and Symphony Halls). ANTI-audio? You been writing too much commercial copy? Re VAC, I have nothing but respect for a small successful company that manufactures great products. I've been self-employed for 20+ years. I have several friends whose small basement ops are now $5-8M music/audio companies, and believe me their noses are to the grindstone. I also just sold a Subie to the wife of the ex-production manager of one of them who, after 15 yrs, got sick of rushing product out the door to meet trade-show dates, brushing aside QA warnings. I wore BOTH those hats in the 70s and 80s at the same time, here in Boston AND in France; I know the stresses implicit when you've promised you'll never ship a product until it's perfect, and then you release a first batch, and then a second level tech finds a flaw two days later....
So yeah, it's easy to say that I got a bad AVATAR, or to incorrectly convince the naive that because I didn't calibrate a mic or measure the drapes my test is flawed, but that's all bull, and the congnoscenti know that. No axe to grind. I'm not into high output impedence tube amps. I tried ONE, May 2000. It went crazy on the VA Fidelio...sounded horrible. Freq resp was crap, as THREE SS amps self-agreed as a group. Here are the numbers: nice and valid. AVATAR since redesigned. Fine. Everybody's happy. 'Cept Lrsky. Why?