• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Audio by Van Alstine ABX Comparator Review

Then they shouldn't attempt to cross religions and use our methods of controlled testing. Their idea was to prove their point using our means as to it being credible. In that sense, we do get to show evidence that could cause corrupt results.
Point taken. But the distortion can't cause corrupt results, it just shows that the thing differentiating the DACs is not distortion. Which I don't think anyone believed it to be. I tried to read up on what people said about A/S about distortion, but it's a 65-page thread and I couldn't figure out a by-thread comment search.

I did find this comment from Josh, though:
I did some basic measurements with Multitone to check whether the box was affecting A differently from B. The FR was untouched. The box seemed to add some distortion, but it was equal between A and B. I mentioned the Multitone 32 numbers in this ASR post, which I screenshotted earlier in this thread.

Once I knew it was equal between A and B, my concern was that any noise added by the box would make it harder to pass the test. Indeed, I didn’t think the DACs sounded as “good” through the box, but I can’t separate that from the fact that ABX tests aren’t exactly as “fun” as normal listening.

So he was never hiding the ball re: distortion. IMO the FR differences are too slight to account for the positive result. So far the only damning thing is the differences in relay sound. And that's just your word against his. For the record I believe you.
 
But the distortion can't cause corrupt results, it just shows that the thing differentiating the DACs is not distortion
And how do you know if the DACs did sound different? From the uncontrolled test and no-data from the single person?
 
And would you say the same about frequency response that depends on volume control setting?
I'm persuaded by Solderdude and others in this thread that the FR changes are de minimis and very likely cannot explain Mound's positive result.

And we do not know how it behaves with change of load impedance and source impedance.
Agreed. I wish that was measured. As well as phase.

To be fair to Amir, the goalposts constantly change. You're never going to convince a dyed-in-the-wool subjectivist. But for me, it's about presenting the best case for the persuadable center.
 
So he was never hiding the ball re: distortion. IMO the FR differences are too slight to account for the positive result.
The order of magnitude was not clear until I did my testing. It is like dipping a spoon in salt an then use it to evaluate two vanilla ice creams. You could argue the same amount of salt is in both samples but the effect is so overwhelming that it would stop any such testing dead in its tracks.

As I noted, if the incredibly low non-linearity in DACs is audible, and impacts such things as sound stage, etc., they the effect of ABX box should have been dramatic. Listener should have surely commented that his beloved music sounded like crap now even though he could distinguish the two DACs. But we did not and instead got comments like this:

"That said, even if the tests confirmed less than I’d hoped, they didn’t contradict what I’d heard before. I never thought DAC A had a shallower soundstage than DAC B in unblinded, level-matched listening, only to find the reverse under ABX."

Surely all that is in the chain right now would destroy any sense of sound stage by subjectivists. But it didn't.

We also have to add this to extreme reluctance to loan me the comparator for testing. None of this makes sense.
 
In over a hundred tests Josh could hear definite differences between DACS, describing them as subtle changes in various areas.

Yet in all those tests he didn’t hear the clear and obvious difference between the relay switch from A to B and back.

And he didn’t notice the significant distortions produced by the comparator.

At the very least this renders his test results null and void.
 
If someone argues that Josh’ results still validly picked out differences between dacs, they would have to declare that even in the presence of 50db distortion/noise degradation introduced by the test box…the DAC differences are not just audible, but easily audible.

I can’t imagine anyone would even try to say that’s credible.

Meanwhile Occam’s razor is lying atop the relays.
 
He hears relays, which is subjective report, but didn't run ABX procedure, which would be an objective evidence. Or have I missed something?
What do you mean by ABX procedure?
Amir reported he and his wife can differentiate the sound. What more can he report you’ll trust? I suppose nothing.
 
And how do you know if the DACs did sound different? From the uncontrolled test and no-data from the single person?
I'm not sure if I understand the import of the question, but I'll answer it literally.

My assumption both before and after Amir's measurements is that the DACs did not sound different. My intuition is that Josh is not intentionally prevaricating, but as Feynman said the easiest person to fool is yourself. And my intuition could well be wrong: he might be lying!

Was the test uncontrolled? The ABX comparator was the control. Just not a good control. Was no data supplied? Looking at the initial article and the update, I see tons of data. GIGO, for sure, the presence of data is no guarantee of anything.

Single person? Theoretically not a problem. If a single person can hear differences, then the devices are not audibly transparent to all people. Of course having a single person poses serious methodological challenges! As with the Goldensound filter test, even under the best case scenario the results should be viewed as inviting further study rather than being dispositive.
 
Last edited:
What do you mean by ABX procedure?
Use AVA ABX Comparator without any device connected. Listen to relays and decide if X is A or B. If I remember correctly, full procedure consist of 10 tests. Then publish the result.
 
Use AVA ABX Comparator without any device connected. Listen to relays and decide if X is A or B. If I remember correctly, full procedure consist of 10 tests. Then publish the result.
Back to square one: why? But no need to reply.
Simpler option: I suggest you buy back the device from Amir and run that test.
 
Back to square one: why? But no need to reply.
Simpler option: I suggest you buy back the device from Amir and run that test.
Sorry if my posts have offended you, I don't try to discredit Amir's work. I try to remain at objectivists position.
 
Dear @amirm
Thank you for running the tests. I am one of the poor suckers that contributed $300.- towards procuring the device only to learn that the difference is the click of a relais!
1791054488097.png

This is the FischerTechnik relais I played with as a nine year old... still have that tick-tock sound in my ears and always knew if it was pulling up or releasing. How I miss those days and those sounds! Just for my fond memories: would it be possible to record the sound, maybe with the microphone you use to make your fantastic videos?

Thank you for your kind consideration!

Kind greetings from Switzerland :)
 
Back to square one: why? But no need to reply.
Simpler option: I suggest you buy back the device from Amir and run that test.
No need to buy it, surely he'd loan it to a member, since he thinks it outrageous that Josh wouldn't loan it to him.
 
I detect that some think this issue still somehow ‘open’.

But it isn’t.

1] There cannot be any performance difference (between dacs under test) that caused Josh to ‘perform so well’ during his testing…his metrics show a very definitive result, but the box he used masked any possible performance differences in the dacs by FIFTY DECIBELS(!!).

2] So then, what other thing was the ‘tell’. Relays are obvious. But more importantly, it was not the DAC performance. That case is closed.

It’s over.

Nothing more need be considered.
 
In over a hundred tests Josh could hear definite differences between DACS, describing them as subtle changes in various areas.

Yet in all those tests he didn’t hear the clear and obvious difference between the relay switch from A to B and back.

And he didn’t notice the significant distortions produced by the comparator.

At the very least this renders his test results null and void.
To be fair he did claim to notice the comparator's distortion, but to Amir's point did not seem to think it was problematic for detecting what must be much smaller differences in the sound. I find this pretty typical of subjectivist testing, where they will listen for changes in noise level or "soundstage" with no regard for the noise level of anything else downstream of the component (like a power cable.)
 
Maybe @mcdn should reconsider and adopt my suggestion for reed relays in the new comparator? :)
Why on earth the designer didn't use them in this one?

Reed relays sound like a faint *tint* if you're 3 cm close and they last forever!
No way you can hear them in the case.

Review was fun, the DUT not so much. A lot could be done better not to affect both the results and most importantly the connected gear under test.
We give the stick to gear with high output impedance and low input impedance and this one goes off the charts, tubes has lower output impedance than this.

And for what? For the use of the passive attenuator? Who adds distortion in the '70's?

Thanks Amir and the rest who payed for this thing to be tested!
 
I have the top off: it has the same number of relays as posted:

index.php


In my casual look, it doesn't look like anything is duplicated. Someone less lazy than me should count. :)
To use this trick one does not need to duplicate relays.
Just a set of DPCO relays where one is energized when the other is not. When changing the A to B the heard sound would be the same.
switching A to B would make 2 different clicks at the same time and B to A would also.

I guess you are saying that this isn't the case.

With the finger on a relay one can feel which one clicks when switching over.
 
Back
Top Bottom