• 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

Which Matrix model are you using? My memory is a little fuzzy, but I think in the other thread we saw some Matrix models are inverted polarity but the model Josh used preserves polarity.
Amir has the original one tested at ASR and the same that Josh used. The issue was RCA vs balanced put and why Matrix added extra opamps at RCA output in later models. This is indicating a fix for polarity. But Josh showed a picture indicating correct polarity. What Amir said above stirs the pot and needs some more data. Whiich unit is inverting? Matrix or Topping. And what output?
 
Whiich unit is inverting? Matrix or Topping. And what output?
The output is RCA as that is the only thing that AVA accepts. I found this out late at night and didn't have time or motivation then to figure out which is out of phase.
 
The output is RCA as that is the only thing that AVA accepts. I found this out late at night and didn't have time or motivation then to figure out which is out of phase.
This was my suspicion from start since it was the matrix model that came out with 16/16 in all pairwisse tests. Also, the notes about timbre and soundstage which is typical notes for those sensitive to absolute polarity. If you find this model inverting polarity, Josh must be wrong. And I start to feel a bit sadness if true. So much work, blood and sweat in this story.
 
This was my suspicion from start since it was the matrix model that came out with 16/16 in all pairwisse tests. Also, the notes about timbre and soundstage which is typical notes for those sensitive to absolute polarity. If you find this model inverting polarity, Josh must be wrong. And I start to feel a bit sadness if true. So much work, blood and sweat in this story.
I would not assume Josh is wrong. I can't imagine getting it wrong if he looked at it like he did.
 
I would not assume Josh is wrong. I can't imagine getting it wrong if he looked at it like he did.
Well I don’t know how many doors this case have had that has been opened and closed. I hope that it is not the Matrix inverting polarity due to all work that has been put in this project. I hope Amir can confirm which unit it is.
 
Had a bit more time to investigate the phase issue. Here is the outcome of comparing 3 DACs:
AVA ABX Comparator blind testing Audio phase difference Measurement.png


Focusing on the red graph, Topping D70s vs Matrix MAQ, it is showing a constant latency of 49 usec vs the other (a bit more than 2 audio samples at 44.1 kHz).

Comparing to Topping D70s this time against a (newly released) Loxjie D30 shows what looks like random phase differential. In reality there is now a 18 millisecond delay (1000 times longer). The sweep doesn't have enough points to capture it so shows that chopped up output. To see that more clearly, here is the unwrapped phase of the same, this time measured with a CHIRP signal:
AVA ABX Comparator blind testing Audio phase difference unwrapped Measurement.png


The massive phase differential due to the very large latency of the one DAC vs the other.

The above of course varies by not only filter selection but filter selection for the DAC chip used. So no assumption can be made about the two signals being in sync even if fed with the identical S/PDIF stream as was the case here.

There are so many traps here that we could test for weeks and still discover more issues with this kind of test setup.

What is the solution? Trust the psychoacoustics. Trust the measurements. When they two combined scream transparency to the source, that is the truth.

If you are going to perform this type of test, best to start with full suite of measurements, verification and much better ABX box than the AVA. Anything else is problematic especially in the hands of a non-technical person.
 
Interesting indeed (incl. the related post #619).

This would mean that.
1: the -40dB A-B crosstalk could potentially be audible (would require AB testing) and could potentially create a different sound in A and B because of the delay.
2: A timing difference artifact, which would be short duration during the switching moment if there is no short mute, would only be audible (and possible tell) during the switching.

@amirm , I still have one burning question that does not involve a lot of hassle and clears things up.
I hope you would be willing to grant me that 'wish' as it would say a lot about the attenuator (distortion and AC behavior is already known)
No APX555 needed, just powered on and multimeter with Ohm setting.

Involves measuring resistance in and 10 measurements in total.
Would like to know the input resistance of the AB box (RCA input), output resistance RCA out and through-resistance (pin RCA-in to pin RCA-out) in 3 conditions.
Requires the box to be powered and in adjustment mode.

  1. input A resistance (tip-sleeve) in 99 setting no load
  2. input A resistance (tip-sleeve) in 90 setting no load
  3. input A resistance (tip-sleeve) in 99 setting with say ... a 10k resistor on the RCA out (as a load, must be resistance)
  4. input A resistance (tip-sleeve) in 99 setting with 10k resistor on the RCA out
  5. output A resistance (tip-sleeve) in 99 setting nothing connected to the RCA in (A selected)
  6. output A resistance (tip-sleeve) in 90 setting nothing connected to the RCA in
  7. output A resistance (tip-sleeve) in 99 setting with shorted RCA-in
  8. output A resistance (tip-sleeve) in 90 setting with shorted RCA-in
  9. resistance between input A-pin and output pin (A selected) in setting 99
  10. resistance between input A-pin and output pin (A selected) in setting 90
 
  1. input A resistance (tip-sleeve) in 99 setting no load
  2. input A resistance (tip-sleeve) in 90 setting no load
I spent quite a bit of time characterizing the input impedance last night. It essentially stays constant in the order of 20 kOhm. I did that as a matter of fact with both AP and my multimeter. Results were similar so I didn't post any. There may be some variability but right now, I have lost patience and interest.
 
I spent quite a bit of time characterizing the input impedance last night. It essentially stays constant in the order of 20 kOhm. I did that as a matter of fact with both AP and my multimeter. Results were similar so I didn't post any. There may be some variability but right now, I have lost patience and interest.

Yes, that means the AD5262 is the 20k version, as expected.
That does not explain the measured (and discrepancy with the dashboard calculations) of the output resistance (basically the wiper).
In the 99 setting the output R would basically be just under 100 ohm (+ source resistance) and in 90 setting around 2k.
These numbers differ a lot from what you measured and what I calculated hence the request.

I understand you are done with it though. :)
 
Last edited:
That is one of the reasons why instant switching between the channels is a wrong method. I have been saying it all the time.
But it was not polarity inverted. The 50 µs latency is not anything or? You would need ms.
 
But it was not polarity inverted. The 50 µs latency is not anything or? You would need ms.
I am speaking neither about polarity inversion, nor 50us or 300us latency. I have been speaking about all possible signal non-integrities related to instant switching (without fade-out and fade-in), should they be residual DC or time delay related. My conviction is based on hundreds such experiments that I performed.
 
Yes, that means the AD5262 is the 20k version, as expected.
That does not explain the measured (and discrepancy with the dashboard calculations) of the output resistance (basically the wiper).
In the 99 setting the output R would basically be just under 100 ohm (+ source resistance) and in 90 setting around 2k.
These numbers differ a lot from what you measured and what I calculated hence the request.

I understand you are done with it though. :)
Looks like AD5262 is operated in shunt mode with a fixed upper resistor
 
In that case the input R would vary depending on attenuation settings which it doesn't acc. to Amir's measurements.
Also the steps appear to be 0.1dB (3 'steps' on the attenuator) which would require complex setting in that case.
I suspect, however that in series with the wiper there is an additional resistor.
 
I've checked the numbers for some DACs from Josh's line-up:

DACChipFilter group delayDelay @ 96 kHz
Matrix Sabre Pro MQAES9038PRO35/fs364.6 µs
Gustard X30ES9039SPRO33.8/fs352.1 µs
SMSL SU-1AK4493S29.2/fs304.2 µs

All numbers are for the linear phase fast roll-off filter. For other DACs, the chip is either unclear (RME ADI-2 DAC FS, three possible versions) or custom and lacks the necessary datasheet (Topping D900, Schiit Yggsdrasil LIM). For the ES9039SPRO, the simulated filter latency is also listed in the datasheet and amounts to 771 µs @ 48 kHz, when group delay would only be 704.2 µs.

I was not sure at first if group delay would be the right metric in this case. However, at least AKM defines very clearly in the AK4493S datasheet that their group delay numbers refer to the delay between the arriving digital signal and the corresponding analog output. But I'm not sure if this is consistent with what ESS uses.

The typical order of magnitude for the temporal resolution of human hearing is around 10 µs, as far as I am aware. However, the stimuli used in such tests will differ from what we are dealing with here.

Filter delays are not the only delays though. There will be other buffers - for example those used as part of clock recovery.
 
Concerning levels that may be correct, but the signals are very unlikely to be perfectly time-aligned between the two input devices. Josh claims he used the same fast linear phase filters wherever that was the option which could get some of the pairings close to a perfect alignment, but some of the DACs in his test were custom R2R implementations and so on.

So I assume that this would mean you could get channel A with 1% of B as a slight pre-echo, while channel B would contain 1% of A as a slight post-echo, right?
So this is indeed the situation. I wonder if @Josh83 would try one of the ABX DAC comparisons that @pma made in the other thread, one that does not have passband ripple differences (echoes). That might give some indication if this is the issue, although it wouldn't be perfectly conclusive either way.

Although with something like just two sample difference the echo and main signal are extremely close together, and the passband ripple will have very large frequency period. And I think the ripple does not even exist if there is not both pre echo and post echo happening. (maybe time to call for help from @J.J. !)
So I think it would be more a rolloff issue or a wide dip or bump in passband flatness if there is ripple, not so much an echo in that case (mathematically the same as a very short delay echo). But if relative delay is longer then it could be more like an echo.

another edit: here is the relationship between passbnad ripple and echoes for those interested. You see that pre echo and post echo are created by passband ripple. But if one signal has pre echo and the other has post echo, it seems they both would have the same amplitude spectrum so then this issue is different than passband ripple caused by a filter.

H(f)=1+2αcos⁡(2πfT) [this is the ripple in frequency domain]

which corresponds in the time domain to

h(t)=δ(t)+αδ(t−T)+αδ(t+T)
 
Last edited:
My reading is there are several issues with the ABX box, some/any/all of which might affect the result, all of which are undesirable, and some of which required definite action by Josh to prevent it affecting his results.

If I'm reading the APS thread correctly, Josh claims to have taken the appropriate actions for each of these problems, even though he didn't mention some/all of them prior to them being raised...

Could someone summarise where we actually are? I'm getting lost with all the different possibilities...
 
Honestly, this thread was IMHO "done" with the first evidence of the "comparator" being inferior to devices it was meant to compare.
So, we are essentially in an endless loop...
 
Back
Top Bottom