Audio Research Reference 6 line preamplifier Measurements

Sidebar 3: Measurements

I measured the Audio Research Reference 6's electrical performance with my Audio Precision SYS2722 system (see the January 2008 "As We See It"). In general, this well-engineered preamplifier's measured performance is similar to that of ARC's earlier Reference 5 SE, which Brian Damkroger reviewed in November 2012, although the Ref 6's volume control operates in 103 steps of approximately 0.3dB rather than the Ref 5 SE's 103 steps of 0.95dB. The unity-gain setting was "61" on the front-panel display. The gain with the volume control set to its maximum, "103," was 12.5dB, balanced input to balanced output, and 6.5dB, unbalanced input to unbalanced output. Both sets of inputs preserved absolute polarity (ie, were non-inverting), and the unbalanced inputs offered an input impedance of 37k ohms at low and middle frequencies, dropping inconsequentially to 31.5k ohms at 20kHz. The balanced input impedances were twice the unbalanced values, as expected.

The output impedances at 1kHz and 20kHz were 305 ohms unbalanced and 612 ohms balanced. At 20Hz, the impedances respectively rose to 600 and 1378 ohms, leading to a reduction of low-bass output into the demanding 600 ohm load, and reaching –3dB at 25Hz (fig.1, cyan and magenta traces). Into a more typical load, the output was flat down to the 10Hz limit of this graph (blue and red traces), and was down by just 1.5dB at 200kHz at the other end of the spectrum. This graph was taken with the volume control set to its maximum; note the superb matching between the channels in this graph (the match was equally good at lower volume-control settings). However, as with the Reference 5 SE, the Reference 6's ultrasonic bandwidth decreased at lower settings. Fig.2, for example, was taken at unity gain; the output is now down by 1.5dB at 40kHz, and –0.5dB at the top of the audioband.


Fig.1 Audio Research Reference 6, balanced frequency response with volume control set to maximum gain at 1V, into: 100k ohms (left channel blue, right red), 600 ohms (left cyan, right magenta) (0.5dB/vertical div.).


Fig.2 Audio Research Reference 6, balanced frequency response with volume control set to unity gain into 100k ohms (left channel blue, right red) (0.5dB/vertical div.).

Like the Ref 5, the Ref 6's channel separation was superb, at >100dB below 3kHz, and still 83dB in both directions at 20kHz. The wideband, unweighted signal/noise ratio, ref. 1V output with the input shorted but the volume control set to its maximum of "103," was good, at 80dB, and these figures improved to a superb 100.3dB right and 99.5dB left when I switched an A-weighting filter into circuit. The spectrum of the Ref 5's balanced noise floor while it reproduced a 1kHz tone at 1V into 100k ohms is shown in fig.3. Some very low-level, odd-order harmonics of the AC supply frequency can be seen in both channels, perhaps due to magnetic interference from the AC transformers being picked up by the tubes' steel pins, while a very slight amount of even-order harmonics is present in the right channel (red trace). But all of these spuriae lie below –100dB (0.001%) and are thus inconsequential.


Fig.3 Audio Research Reference 6, spectrum of 1kHz sinewave, DC–1kHz, at 1V into 100k ohms (left channel blue, right red) (linear frequency scale).

Fig.4 plots the THD+noise percentage against the left channel's balanced output voltage into 100k ohms. Below 2V output the distortion is buried beneath the noise floor, indicated by the upward slope of the trace with decreasing voltage. The distortion slowly rises above 2V, but is still just a low 0.1% at 62V RMS, which was when the Audio Precision was outputting its maximum signal of 15V. The Ref 6's balanced output clips into 100k ohms at more than twice the Ref 5's 32V. The single-ended output clips at 12V into 100k ohms (fig.5), which is still well above any level needed to drive a power amplifier into overload, though the distortion at lower levels is higher than it is with the balanced output. The balanced output clipped at 2.7V into 600 ohms; like the Ref 5, the Ref 6 should not be used with power amplifiers having an input impedance of less than 10k ohms.


Fig.4 Audio Research Reference 6, balanced distortion (%) vs 1kHz output voltage into 100k ohms.


Fig.5 Audio Research Reference 6, unbalanced distortion (%) vs 1kHz output voltage into 100k ohms.

I examined how the THD+N percentage varied against frequency at a balanced output level of 2V into 100k ohms, which is both sufficiently high to ensure that I'm measuring THD rather than noise, and close to the maximum that will be required of the preamplifier in practical use. The left channel's distortion (fig.6, blue trace) is slightly higher than the right's (red), but both are extremely low in absolute terms. Fig.7 reveals more second harmonic present in the left channel than in the right, but the distortion is still low. Intermodulation distortion was also extremely low (fig.8).


Fig.6 Audio Research Reference 6, THD+N (%) vs frequency at 2V into 100k ohms (left channel blue, right red) (linear frequency scale).


Fig.7 Audio Research Reference 6, balanced spectrum of 50Hz sinewave, DC–1kHz, at 2V into 10k ohms (left channel red) (linear frequency scale).


Fig.8 Audio Research Reference 6, HF intermodulation spectrum, DC–30kHz, 19+20kHz at 2V into 100k ohms (left channel blue, right red) (linear frequency scale).

The Reference 6 continues Audio Research's tradition of producing well-engineered preamplifiers whose measured performance reveals little hint of the presence of tubes in the circuit. The only criticism I have is that unlike the Ref 5, the Ref 6 dispenses with Audio Research's traditional control knobs.—John Atkinson

Audio Research Corporation
3900 Annapolis Lane N.
Plymouth, MN 55447-5447
(763) 577-9700

A. Hourst's picture

“That may offend the sensibilities of those purists who demand absolute accuracy, but in audio, there's really no such thing”

I don’t know what’s most funny: Michael Fremer who thinks he can “offend the sensibilities” of some objectivists, or the fact that he thinks this will happen from such an empty, predictable commonplace as the “immeasurability” of good sound.
There’s really no such thing, you say, as absolute accuracy in audio. Never mind the fact that the usual tenants of a “good enough” accuracy only ask it to be better than the human ear sensitivity, which is rather easily achieved with modern electronics, Mr Fremer don’t even recognize “absolute accuracy” as an existing horizon in audio. Absolute accuracy conceptually exists in audio as much as it exists in photography or in watchmaking. However, if someone is trying to push a 14 000$ piece of electronic whose performance can be bought for less than 1/10th the price, rising up the confusion by saying things like “in audio, there’s really no such thing as absolute accuracy” can be good practice.
One thing will never happen: a blind ABX comparison of this ARC preamp with a 1000$ similarly measuring one, to put to the test this idea that dollars can get you what measurements and science can’t.

ChrisS's picture

And has never happened, because a blind test in this situation is not practical nor very useful.

ChrisS's picture that you acknowledge that you are the only one who keeps asking for something that will never be done!

Are you the only one who needs to have your idea tested?

Johnnyjajohnny's picture

I believe a lot of people would actually like to see a blind test like A. Hourst suggests, so he's not the only one.
I, and probably many others would find it very interesting, and I'm not convinced it would necessarily turn out one way or the other. So, it would be nice to find out if there really are audible differences between two amps that measure the same :-).

Maybe you have already noticed that the blind test advocates usually ONLY ask Fremer (or anybody else) to do blind tests with certain things.
As far as I know, nobody has ever asked him to do a blind test comparing ANY of the speakers he has ever owned to a pair of 200 dollar computer speakers or even to 1000 dollar KEF speakers. Nor do I see any requests for a blind test between a Lyra cartridge and a 50 dollar Audio Technica cartridge.
And why?
Because the objectivists already know that there are readily audible differences between most speakers and also many cartridges. As for cartridges, I've compared my own cart to 35 others, and the blind tests I have done, only with the ones that were the most similar, I passed. I even passed blind tests on some adjustments (overhang, stylus pressure, etc.) that I didn't expect to pass.
Even some of the very first ABX tests done in the late 70s and early 80s showed that speakers and volume levels were by far the easiest categories to hear a difference. On that note, recently I successfully ABX'ed 0.2 dB volume difference on actual music with 15 out of 16 correct (the test I'm the most proud of).
So, the objectivists are usually (not always, as some objectivists really are obnoxious and OCD) asking for blind tests of more controversial topics.
Personally, I think that blind tests will/should reveal differences between certain amps, but some objectivists disagree, and amps seem to be the category that splits the crowd a lot. But even hardcore objectivist Arny Krueger said in "The great debate" in 2005 that he has successfully heard differences between power amps in blind tests.
Then there are categories where it's simply highly unlikely that people can hear a difference, because nobody else has done that so far in a properly conducted blind test. This category includes hi-res vs. the same material down-sampled to CD quality, analog tape vs. a properly digitized copy, and cables. However, some cables are not transparent in the sense that they can be faulty or have certain properties (impedance, etc.) that will alter the sound. Someone at Hydrogen Audio did successfully ABX speaker cables, and a measurement showed quite a different frequency response for the two pairs. Also, Audio Critic showed in the early 90s that certain cables rolled off prematurely or had a spike in the treble, and some of this is audible, and some people buy these cables for exactly those audible properties.
And lastly there's the category of products that simply cannot produce any audible difference based on the currently known laws of physics. This includes expensive power cords and tiny acoustic products by Synergistic Research. Although some people report differences, so far no properly conducted blind tests have shown any difference – quite the opposite.
It's also worth noting that Synergistic Research, Nordost and Audioquest have been shown to do fraudulent demos, where they changed the volume level and used other tricks to "show" an obvious difference that everybody could hear. Obviously, they do this as their products simply don't produce a difference.
This is why we need blind tests. Blind tests are the kryptonite for the golden eared, but also their bragging rights if they pass (I've passed many and failed many) :-).

ChrisS's picture

See John Atkinson's many, many words that he (and others) has written about blind testing and you'll understand why blind testing will not be done by Stereophile any time soon.

Bottom line, Double Blind Testing is neither practical, nor particularly meaningful when evaluating audio equipment.

Bottom line #2, no one in the entire audio industry does DBT.

On the other hand, "Single Blind" testing, even when done informally, can be fun and informative, but the results can no way be considered "scientific".

Johnnyjajohnny's picture

Everybody at Stereophile opposes double blind testing because the tests don't give them the results they want. As simple as that.
I will be blunt here and say that saying ABX tests are invalid is downright stupid.
The writings of Atkinson, Fremer et al exemplifies exactly what the problem is with the audio press: The most scientific tool (double blind testing), which is used in every other form of scientific testing of any kind, from testing medicine to people with paranormal abilities, is by the audio press deemed "useless" and "unscientific" simply because it doesn't corroborate the findings that the critics found when they could see the name tag and knew the price.
Richard Dawkins made an excellent two-part programme called "Enemies of reason" (it's on Youtube), and in the first part he speaks to an astrologer, who gives exactly the same reasons as the audiophiles, "I just don't believe in the experiment", when Dawkins suggest they give out horoscopes to random people. He also says "If your intentions are mischief, what you get back is mischief.". Fremer said: "If the test is stupid, what you get back is stupid". Same thing.
As Dawkins remarks: "I thought you would be keen to try it out if you're so sure that your horoscopes are accurate, so that makes me think that in your heart of hearts you don't belive it. I don't think you're prepared to put your reputation on the line."
When Dawkins visits a double blind test of dowsers done by someone else, they have the same excuses when they can't find water.

I've had four phono preamps in my possesion or on loan in the last couple of years. I made level matched recordings of all them, and I succesfully ABX'ed all of them. I can easily post my logs, 'cause unlike the astrologer or the staff at Stereophile I have no problem putting my reputation on the line. I haven't measured the preamps, but I think they simply have slightly different frequency responses. What the explanation is doesn't matter to me - all that matters is that I could tell them apart, usually with 15 out of 16 correct. This just goes to show that ABX tests work. If you have one minute to spend, you can verify it for yourself: Download Foobar and its separate ABX plugin. Load an Iron Maiden song as A, and an AC/DC song as B (or whatever you prefer) and then ABX them. You will have 20 out of 20 correct in less than one minute.

Bottom line, Double Blind Testing is practical, easy and very meaningful when evaluating audio equipment. All you need is a switch (all mine were done in Foobar) and some time to spend, which you would have when evaluating anyway.

Bottom line #2, no one in the entire audio industry does DBT, because then many companies would go out of business. Sure, many would remain in business and rightfully so, while others would go out of business. Some companies really do deserve to close their doors, like the scam companies like Synergistic Research and Nordost (an elaborate blind test was done with power cords from Nordost: An enormous fail) that charge ridiculous amounts of money for placebo effects and then threaten to sue when the thruth is exposed.

As for Audio Research, I know that some people love them, others despise them. I've only listened to it once, which was really just casual listening of speakers, and I have no quarrel with them.

ChrisS's picture

Please actually read the articles by John et al. And take a course on scientific research and how to set up and implement a scientific experiment.

What you express is opinion and what you describe is not science.

Allen Fant's picture

Great review- MF.
until I can demo the new Ref6, I feel the Ref5SE, is still the best tubed pre-amp in the ARC arsenal.'s picture

Absolutely well written. Yes the Reference 6 is all that and more. Even the new Foundation Series LS28 betters the Reference 5 SE now.
Yes Casework and other improvements have trickled down to the Foundation series.

Vade Forrester's picture

This was one of the best reviews I've ever read. Well done, Michael.

Vade Forrester
Reviewer, SoundStage! Network and The Absolute Sound
My words=my thoughts.

WJ ARMSTRONG's picture

I agree with my fellow reviewer Vade Forrester - this was unusually entertaining and successfully conveyed some quite subtle notions with a lovely light-touch clarity. Almost as enjoyable as listening indeed!
Thanks Michael.
Bill Armstrong - 6moons

jsch123's picture

I own it and it's lovely. The best preamp I've ever owned and it's really brought me to a point of finality . I mean, not really because I'll be tinkering for the rest of my life, but it could easily be finality. If that's even a word.

I agree with just about everything MF said. Especially the "a lot of meat on its bones". But you know it remains exceptionally open and transparent and dynamic at the same time. It's just lovely. Gone is ALL the grain. Smokes both my VAC preamps.