If Either of These Amplifiers is RIGHT . . .

It was January 1986. Stereophile's then publisher, Larry Archibald, and I were driving in his diesel 'Benz sedan from Las Vegas to Santa Fe. We had shaken hands at the just-concluded Consumer Electronics Show on my replacing J. Gordon Holt as the magazine's editor, and now, during the 750-mile drive, we mapped out the strategy to take what was then an "underground," digest-format, somewhat irregularly published magazine to the position of dominance in audio publishing it still enjoys.

As Larry and I discussed on that 13-hour drive from Nevada to New Mexico, we began publishing every month with our October 1987 issue. We had planned to change to the large format you hold in your hands in January 1993. However, we missed that target, not hitting it until 25 years ago, with our January 1994 issue (footnote 1). And in January 2019 as in January 1994, our cover features two amplifiers. One of them is the 25th Anniversary "Reference Series" version of Cary Audio's CAD-805 single-ended triode monoblock, that originally appeared on that January 1994 cover and is reviewed in this issue by Art Dudley. The other is Cambridge Audio's Edge A solid-state integrated amplifier, also reviewed in this issue by Ken Micallef. While the Cary was joined on the 1994 cover by a solid-state Krell KSA-300S, the headline a younger me wrote back then still applies: "If either of these amplifiers is RIGHT . . . the other must be WRONG."

One of the things Larry and I talked about in our 1986 odyssey was the inclusion in the magazine's equipment reviews of measurements, which began in fall 1989. As I wrote a few years ago, my model for a review of an audio product was, then and now, a review of an Ortofon phono cartridge in the July 1966 issue of the English magazine Hi-Fi News, written by the magazine's then editor, John Crabbe. The measurements in that review and the descriptions of the sound quality went together; one could not be read without reference to the other, and vice versa. Each supported the other.

But with the amplifiers on our January 1994 and 2019 covers, this model breaks down. Our reviewers praised both the Cary's and the Cambridge's sound qualities, but when it comes to these amplifiers' measured performances, things could not be more different.

The large, heavy, inefficient, single-channel Cary is intolerant of low impedances, suffers from hum, and delivers only a limited amount of power, with a level of distortion that the late John Crabbe, for whom I worked at Hi-Fi News from 1976 to 1982, would have dismissed with a snort. The two-channel Cambridge delivers respectably high power even into low impedances, with vanishingly low levels of distortion and noise. And, as befits a 21st-century amplifier, it offers a full array of digital inputs, including a USB port and a Bluetooth antenna. These models represent polar opposites in amplifier design, but they can't both be right. Only one can be telling the truth—the other must be lying.

Similarly, Chord's Qutest DAC, which I review in this issue, has state-of-the-digital-art measured performance, while the identically priced BorderPatrol DAC SE, which Herb Reichert and Jon Iverson reviewed in our September and November 2018 issues, measured so poorly that I was concerned that our review sample might have been broken. It wasn't, and, as Jon Iverson wrote in his Follow-Up, the lies the BorderPatrol DAC told were lies of commission: "It purred like a sweet, sultry voice, softly caressing my ears, nibbling them gently, even as it lied to me with every word. Contrast that with what I prefer to hear: a calm, even voice telling me the unvarnished truth, even if that truth may hurt a little."

This in turn led to reader Duane Jackson writing, with more than a tinge of irony, in this issue's "Letters," Jon Iverson "has saved me from my mistaken enjoyment . . . The beautiful sound of strings that I hear at live concerts is a mistake made by my brain analyzing signals from my ears. Those strings should sound steely."

The question is: Does that "mistaken enjoyment" stem from what the component is doing wrong, or despite it? I know how Herb Reichert's "Mr. O" would respond. As for me, 30 years ago, I announced Stereophile's introduction of measurements by saying "Measurements . . . will lift us above the level of the primitive. Reproduced music will not then be magic; it will just sound magic." But as we approach the 21st century's third decade, the arguments over what sounds musically magic and what measures well continues unabated.—John Atkinson

Footnote 1: Click here for a timeline of Stereophile's first 40 years.

rt66indierock's picture

This is just a people problem. Everyone is wrong. A great way to kick off 2019 the year to criticize the reviewer.

"If it measures bad and sounds good, maybe your hearing isn't as good as you think."

Solarophile's picture

I think that quote is the most reasonable and right way to think about the difference between the subjective reviewer's comments and objective results most of the time. I don't know why it doesn't seem obvious to many audiophiles. It's almost like some seem to think that those golden ears are infallible.

Jack L's picture


Sorry I totally disagree to yr above statement.

Read this, pal:-
"I no longer regard as fruitcakes people who say they can hear something & I can't measure it - there may something THERE ! " quoted Richard Heyer.

So believe what we perceive in our brain when we hear the difference.

Measurement methodology todate is far far behind what our brain can perceive & comes up with data irrelevant to what we hear, unfortunately.

Lets's take a realistic example: how come a vacuum tube amp which measured 5% total harmonic distortion sound so much better than an solid state amp which measured 0.00005% total harmonic distortion using same measurement equipment?

This was taken from a Master Degree thesis paper some 10 years back on
objective measurement methodology relevant to our aural perception.

Listening, but not measurement, is believing !!!!

Jack L.

HammerSandwich's picture

I've never read Crabbe, JA, and wish I could see his review of the Borderpatrol DAC. Do you believe that Crabbe could have made sense of this DAC's results, both sub- & objective? If so, would HR's or JI's review hew closer to Crabbe's approach, or are we talking about 3 distinct styles?

Bogolu Haranath's picture

Hey, what can I say? ..... I'm just a ..........

"Desperate Man" ............ Eric Church :-) ...........

tonykaz's picture

It's consistant with all things Human that people can Love one type more than another, isn't it?

I'm a fan of Per Abrahamssen's Electrocompaniet Amplification and NOT a fan of Conrad-Johnson ( for whom I was a Full Range Dealer ). Many of my Thiel Audio customers demanded Tubes so I carried their favorite Brands .

Silly of me to think of them as being wrong or as having tin ears but I did. Sorry

Now I realize that everyone hears differently and prefers different types of Audiophile performance.

If they seem wrong it's based on MY prefrences so I have to be careful to not condemn my neighbor's love for Bose 3s & Pioneer Receivers for gods sake .

Reading about people working with their favorite Audio stuff is what makes Stereophile amazing and worthwhile. Where else could we find a Kal Rubinson explaining 4+1 and 5+1 in a meaningful Audiophile way ?

If loving you is wrong, I don't want to be right!

Tony in Michigan

supamark's picture

printing the letter writer's email address online - you've just increased the amount of spam he receives at that address by like an order of magnitude.

Ortofan's picture

... the manufacturer's comment by David Hafler following the review of his XL-280 power amp:

Then, ask yourself what you expect of a power amp - or of a high-fidelity sound reproduction system, in general. Should the equipment be as objectively accurate as possible by imposing as little as possible of its own character upon the signal passing through it? Or, should the equipment make the resultant sound quality in some manner more subjectively pleasant by adding some distortion or noise to and/or modifying the frequency response of the signal passing through it?

Essentially, do you really want the truth or would you rather hear lies?

A question for JA is why the Hafler SWDT was never incorporated into Stereophile's test suite?

The XL-280 was claimed to be the most accurate amplifier available, and I don't ever recall any other manufacturer challenging that assertion. To cite one product, presumably the Benchmark amp - due to its extremely low levels of noise and distortion - would achieve a better score on the SWDT, and it would be interesting to know exactly how much better it might perform.

michaelavorgna's picture

...they are not performing a job nor are they evaluating the quality of a recording. While these things are true for some, the notion that one can be lied to when listening to recorded music for pleasure makes as much sense as believing a photograph is a lie because it can't sing.

Michael Lavorgna
Editor/Publisher, TwitteringMachines.com

CG's picture

Imagine a sex machine. By design, it responds to brain wave patterns and does whatever it needs to maximize the stimulus required for the desired response.

Is that better than a loving companion? Is one right and the other wrong? Or, as Michael suggests, is that even a valid question?

Perhaps more to the point, over the decades test equipment has progressed so that the standard measurements of noise and harmonic distortion are really sensitive, probably far beyond what the inventors of those tests ever dreamed of.

But, does that make the tests a complete or suitable evaluation tool of the objective performance of an audio reproduction device? Other technical areas that use linear amplification moved on from that approach decades ago.

Actual music and other reproduced sound is a whole spectrum of tones that create IMD in imperfect amplification and other signal processing devices. In addition, that spectrum changes content and relative phase quickly over time. Modern audio test equipment does not allow for any of that. Instead, it achieves sensitivity by averaging the same tones over a long period of samples. Is that valid? (It isn't in the telecommunications biz.)

In addition, virtually all these tests used today in the audio world make almost no attempt to evaluate the performance of the electronics within a system. In a perverse way, that has the effect of minimizing the importance of that performance aspect in the design as well. Yet, nobody uses an amplifier by itself - how would that work? Anybody else see a problem with that?

This is about like a physician taking your temperature and blood pressure and then declaring that you are healthy or not, based on just those two tests.

volvic's picture

Love the info on your site, very informative. Glad you're back.

grantray's picture
Ortofan's picture

... Michael Lavorgna and Jon Iverson on the topic of whether or not one can be lied to by any of the various components in the playback system when listening to recorded music for pleasure?

michaelavorgna's picture

Here you go - even if 'one can be lied to...', and the notion is patently absurd, it wouldn't matter in the least.

End / debate.

Michael Lavorgna

Ortofan's picture

... lying to you. Tell him that his "notion is patently absurd" and see how long you keep working at Stereophile.
Oh, wait ... you're not.

michaelavorgna's picture

...perfect sense.


Michael Lavorgna

ok's picture

..with pure pleasure indeed – and quite safe a pleasure no doubt!

ok's picture

1st man: [..] that's all.
Judge: Sounds right.
2nd man: Enough, now listen [..]
Judge: Sounds right.
3rd man: Wait a minute.. both right?
Judge: Hmm.. sounds right.

Bogolu Haranath's picture

"I never did give anybody Hell ........ I just told the truth and they thought, it was Hell" ........ Harry S. Truman :-) ...........

Bogolu Haranath's picture

"There are no whole truths .... All truths are half-truths ..... It is trying to treat them as whole truths that plays to the Devil" ......... Alfred North Whitehead :-) ..........

Bogolu Haranath's picture

"Contradiction is not a sign of falsity, nor the lack of contradiction a sign of truth" ........... Blaise Pascal :-) .........

Bogolu Haranath's picture

"The pendulum of the mind alternates between sense and nonsense, not between right and wrong" ........ Carl Jung :-) ............

Bogolu Haranath's picture

"I love the gray area between right and wrong" ......... Dan Brown (Author) :-) .........

Long-time listener's picture


Ortofan's picture

... he contrasts the "steely" sound of string instruments as reproduced by his system at home with the "beautiful" sound heard in a concert hall setting is the relative difference in distance from his ears to the instruments in the concert hall versus the distance from the microphones to the instruments for those recordings that seem to exhibit an unpleasant sound quality.

Russell Dawkins's picture

Is that many circular diaphragm condenser microphones have breakup modes in the high frequencies contributing to this steely sound, but also any good engineer compensates for this in the mixing process, according to his/her judgement. A bigger factor, also related, is the wide disparity in the representation of bass frequencies between even pro grade monitoring speakers not to speak of home speakers. Too many of both types 'boom', leading to a common under-emphasis of bass frequencies in recordings intended to be realistic, which exacerbates the over-emphasied highs.
Here's a reasonably good string sound:

Robin Landseadel's picture

They're both wrong.
Hope that solves it.

PS: A violin heard from a meter's worth of distance is always steely.

ok's picture

..but cold not whatsoever.

Robin Landseadel's picture

String sound can be cold. Remember that room acoustics are a big factor in the tone color of an instrument. And some players just sound nasty. That's reality. That's what recording engineers have to work around.

Bogolu Haranath's picture

Choose your poison? :-) .......

Bogolu Haranath's picture

Steely Da(m)n :-) ............

Bogolu Haranath's picture

"Am I Wrong"? ........... Nico & Vinz :-) ...........

Wimbo's picture

I don't care what the color is.

Golden Ears's picture

Audio truth.

If the reproduced musical event were indistinguishable from the original, then you can say you have it “right”.

Music is an emotional experience. That is what we are after, to feel moved. Different instruments of different qualities help increase our musical enjoyment. Depending on the composition, they don’t have to always be the finest quality (“more cowbell!”) but they are selected for a reason.

To audiophiles, typically better sonic reproduction of the aspects that mean most to us gives use a greater emotional impact.

Everything in the recording chain, The microphones, cables, tubes. Mixers, recording medium , mixing editing, mastering, storage medium, playback device, amplification, cables, speakers, room and frankly whether or not you cleaned your ears recently has an effect on what we hear versus the actual original event.

Everything is a filter.

Not all filters are the same or have the same impact. We all know you can hobble together a system where everything measures well but sounds “too dark and plodding” or “ too bright and fast” . Indeed System balance is important and should compliment things like the limitations of your room.

So is more of the truth of the recording a good thing always? Not necessarily, for instance I have been at audio stores and heard the salesman say

“ There, did you hear that, did you hear the chair squeak in the far rear of the audience? These speakers?amps/cables have incredible resolution“.

Recording engineers go to great lengths in microphone placement to avoid highlighting chair squeaks, audience sneezes, coughs, cell phone buzzing etc. in mix downs you try to downplay these, and same cases in every part of the recording and mastering chain. So why should we try to bring these annoyances back to the forefront? It is “audio truth” but it is not part of the emotional experience. Far better for a salesperson to point out how emotionally moved you are by a speaker/amp/cable “Wow, I can see you have shed a tear”, “Really gets you dancing “ “Wants to make you sing along, right?” etc. Those feeling are also part of audio truth. And in my humble opinion they mean a lot more than the squeak of chair in the blackest of black backgrounds that go to infinity to let you hear the harpist’s quietest farts.

Thats not to say signal to noise isn’t important, I do think that signal to noise ratios of about 110db or better tend to sound better than lower S/N ratios but sometimes a bit of the baby can get thrown out with the bathwater chasing vanishingly low distortion with negative feedback etc.

Everything impart its own filter sound.
DAC’s , the vast majority of them, personally I have heard less than 5 DACs (made from 1990’s tp present ) that sound relatively natural without errors of commission that are gritty and spotty in high frequencies and make instruments sound vaguely out of tune or out of step. I don’t care f you call it jitter, or a steep brick wall filter, or low sampling rates, or the fact that there is thermal noise in the chip limiting us to 21 bits of resolution..whatever you want to call it, I hear something grossly unnatural and unpleasant so the only time I can enjoy digital from a DAC not from that list of those 5 pleasant DACs is when I am being distracted doing a physical activity typically where I’m moving faster than 50mph like driving or snowboarding 30-70mph or skateboarding 25-50mph. I need the mental distraction of a quasi dangerous experience to not have the Digital grunge and lack of emotional connection so apparent .

Some filters tend to add a rich presence to the sound… like tubes, tape, vinyl. And even though the lowly Cassette tape has hiss, and often was played back with dolby phase shift, noticeable wow and flutter (speed variances much larger than digital) the pleasurable filter added by the tape and what it may have left out made tithe most successful selling recorded medium despite Vinyl and Reel to Reel being around. Here was a medium you could listen to while distracted driving, or doing sports, and it had no surface noise save a bit of hiss, plus it was immune to skipping. This was the heyday of enjoying recorded music. And it was the most pathetic recorded analog medium (next to MP3 and mini Disc and frankly if you heard even a prerecorded ferric cassette through a Tandberg 3014a deck you would likely prefer it to most of what is considered the best digital, because most of the emotional content is not nearly as damaged as it is with digital.

Now I also feel that the average playback system acts as a filter which can determine what music becomes popular. The music whose emotional content is LEAST damaged by the jumble of filters in the playback system tends to become popular. So when we had mono tube radios Vocal harmony of mostly midrange was the least damaged. In the 1960s even Keith Richards realized that he had to make his music sound good though a diction type mono tape recorder (he used one to record street fighting man)In the 1970’s rock music that used distortion musically was least damaged by cheap solid state (Hendrix and Led Zeppelin ). Discotheques made disco and 1980s music sound less damaged at high volumes , Radio was popular because The Bozak Broadcast Mixer sounded so good.1990s had digital replacing vinyl and industrial bands like Nine Inch Nails embraced the harsh digital sounds and made it part of the music. In the 2000s iPods with low bit rates and crappy ear buds had groups like Edie Brickell and folk sounding groups emotional content least damaged. The subwoofers tossed into cheap cars without upgrading the front speakers had Rap and hip hop become the least damaged music emotionally. Computer speakers have rouble producing anything other then EDM. And well protools is a filter all to itself.

People vote emotionally for their music. And if their music doesn’t sound good- they will listen to talk radio. Club DJs noticed it early on in the digital era that people didn’t like to dance to Digital as much as with the exact same song on vinyl despite wow and flutter, feedback, and surface noise. And attendance started going down, and as vinyl became replaced by digital most of the nightclubs closed because digital through already glaring distorted PA systems was just too detracting from emotional enjoyment to make it worth going out anymore. Hardly perfect sound forever, more like I don’t like going out and listening to recorded music. EDM helped revive this, but IMHO we will never have the nightclub base playing analog music and rock we once had without analog music without digitization. So many clubs use digital compression and digital EQ, that there is no way to make analog pass through the system without being stripped of a great deal of its emotional impact. Again EDM, Industrial , and Rap Hip Hop and grudge seem to be the only genres that are least affected by digitization. Promoters try to emotionally involve the crowds with more lights and pyrotechnics when they really should be looking at the full audio chain.

So which amplifier is right? The one that gives the listener the greatest emotional connection. Having heard the Krell KSA series (great bass PA amps BTW and fantastic for difficult loads) and the founders edition of the Cary 805 founders edition mono blocks with a few mods and top tier tubes (I have demonstrated with those amps at RMAF and T.H.E. Show Newport Beach I’d have to say the Carys are the more “right” amp as the musical EMOTIONAL connection is stronger. I have done demos with Solid State amps like the Odyssey Stratos and also can get good results but the tubes dollar for dollar tend to be more right in each price class over $2500. Now if I were just listening to EDM, Pearl Jam , NIN through Magnepans Id go with the KSA in a heartbeat, but for the vast majority of wide genres of recorded music (both analog and digital ) the Cary will be the better fit.