Pano,
Good. And this is exactly the thing what Sy had refused to do here. Sticks to one test signal, the most primitive and less revealing, and that's it. When I suggested that maybe a signal better fitting to this test should be applied, he said he is not interested in it.
I know, I know, we will anyway only listen to 1KHz pure tones in our cosy music evenings, so no reason to go further..
Ciao, George
Good. And this is exactly the thing what Sy had refused to do here. Sticks to one test signal, the most primitive and less revealing, and that's it. When I suggested that maybe a signal better fitting to this test should be applied, he said he is not interested in it.
I know, I know, we will anyway only listen to 1KHz pure tones in our cosy music evenings, so no reason to go further..
Ciao, George
@Joseph K
You have a Hiface unit to hand, and attenuators evidently. You have not been backward in presenting screenshots of the digital interface which you insist show the mitigation of jitter by the attenuators. You also have an SPDIF receiver, which implies a DAC.
Why then have you not presented screenshots of FFTs showing the before and after effect on the analog output? I don't know exactly how SY is doing the FFTs, but I get the impression he's capturing the output via an M-Audio card and post-processing it, or it could be real-time. Anyway, I'm sure he can supply the details. If I'm correct then all you are missing is a second card and some software.
Not, of course, that it's any business of mine to tell you how to make your own case.
w
You have a Hiface unit to hand, and attenuators evidently. You have not been backward in presenting screenshots of the digital interface which you insist show the mitigation of jitter by the attenuators. You also have an SPDIF receiver, which implies a DAC.
Why then have you not presented screenshots of FFTs showing the before and after effect on the analog output? I don't know exactly how SY is doing the FFTs, but I get the impression he's capturing the output via an M-Audio card and post-processing it, or it could be real-time. Anyway, I'm sure he can supply the details. If I'm correct then all you are missing is a second card and some software.
Not, of course, that it's any business of mine to tell you how to make your own case.
w
Last edited:
Just as I said in earlier posts. My laptop -> USB/spdif -> DCX2496. Analog out of DCX (sine wave generated by Audiotester) to M-Audio 192 PCI in my lab computer, with gains adjusted to accommodate level differences (pro vs consumer). AudioTester in the lab computer FFTs the sine wave. 64k points, 96k sample rate. Apodized, but I don't recall the function at the moment- if it's important to you, i can always look at it again when I get home.
The simplicity of the waveform makes jitter effects easy to spot with a minimum of confounding effects.
Thanks, SY.
I have a piece of free software here called VA, it has been mentioned on the forum before but it hasn't got a great deal of attention, which I think is a shame, because it looks good to me. It has oscilloscope functions and spectrum analysis. I have written to the author to ask if it is possible to run the spectrum analysis on a file, and in particular if a number of FFTs could be averaged. Whatever the answer, I am encouraged to note that I could run the real-time test myself, as could most people here, my best A/D is an M-Audio 2496 for the moment, but this could change.
w
I have a piece of free software here called VA, it has been mentioned on the forum before but it hasn't got a great deal of attention, which I think is a shame, because it looks good to me. It has oscilloscope functions and spectrum analysis. I have written to the author to ask if it is possible to run the spectrum analysis on a file, and in particular if a number of FFTs could be averaged. Whatever the answer, I am encouraged to note that I could run the real-time test myself, as could most people here, my best A/D is an M-Audio 2496 for the moment, but this could change.
w
Yes, it's very easy to do, which is why I was surprised that purveyors of these expensive USB devices don't show the effect of their gizmos on the output of the signal chain- unless, as seems to be the case, there's no effect at any kind of practical level.
The M-Audio 2496 is a pretty good card. Use it in conjunction witha good analog interface (like Pete Millet's or the one I use) and you have a very powerful tool. I get pretty much identical results with other pieces of software (e.g., ARTA, HOLM) but use Audio tester just because I know where all the buttons are.
How?
At this stage I'm assuming that you have listened to both devices & you can't hear a difference, otherwise continuing in this vein would seem disingenuous. This indicates to me that you are doing something in your playback that masks any effects.
I see that you are using the DCX even after the criticisms laid at it's door by many! Did you try the other DAC? Are you intending to?
No, you were clear that I should stop. And given your reaction to results you didn't like, I'm just waiting for instruction about packing it up and sending it to whomever you think WILL give you the results you want.
I'm not comfortable with being given a specific target for results, then try to design some sort of test to give the pre-determined outcome; for me, that's not honest, though other people can certainly have different views.
Really, SY, you have invalidated the measurements of the attenuators which I claimed would show how they cut down reflections - you set up the experiment so that there were no reflections & then refused to alter it.
In testing the devices you used a DAC which was said to unsuitable on a number of fronts:
- it used an ASRC
- it analogue output stage was criticised & damned
You suggested you would be using another DAC which you didn't.
You now are using various excuses to avoid validating or defending these tests.
And you say I'm looking for pre-determined outcomes - the determination is all on your side, I'm afraid.
I believe that you can't hear any difference & so you are dogged in your claims. On my behalf I have heard significant & unmistakeable differences between the modified Hiface & the stock version. I'm sure that your $30 Chinese device is no better that the stock Hiface & I would venture to say, a lot worse, knowing the chip-sets involved.
You must think that I'm either a charlatan or a fool who would submit a device for testing that I'm not sure is definitely better than a run-of-the-mill USB-SPDIF device?
As I said, I don't distrust your belief in your results but your unwillingness to investigate belies an agenda, I'm afraid.
In testing the devices you used a DAC which was said to unsuitable on a number of fronts:
- it used an ASRC
- it analogue output stage was criticised & damned
You suggested you would be using another DAC which you didn't.
You now are using various excuses to avoid validating or defending these tests.
And you say I'm looking for pre-determined outcomes - the determination is all on your side, I'm afraid.
I believe that you can't hear any difference & so you are dogged in your claims. On my behalf I have heard significant & unmistakeable differences between the modified Hiface & the stock version. I'm sure that your $30 Chinese device is no better that the stock Hiface & I would venture to say, a lot worse, knowing the chip-sets involved.
You must think that I'm either a charlatan or a fool who would submit a device for testing that I'm not sure is definitely better than a run-of-the-mill USB-SPDIF device?
As I said, I don't distrust your belief in your results but your unwillingness to investigate belies an agenda, I'm afraid.
Yes, my diabolical use of normal lengths of off the shelf coax and a correct termination.
I thought that argument might be coming. In the case of jitter, a single tone is legit, even desirable. It will clearly show the jitter if it's there. It would be very hard to see in the FFT of music!
The only thing I can add is that the top octave 10K-20K seems to be where the jitter mavens like to test. It seems to show up better there for some reason. That's the only thing I'd add to the test.
Yes, my diabolical use of normal lengths of off the shelf coax and a correct termination.
So you ignore my premise of how the attenuators work & decided to show that you could terminate correctly (with 72R mind you!) - bravo & that proves what?
If you are going to obfuscate the issues why did you agree to do these tests?
Please remember (this is the fourth time I've said this, in case you wonder why I don't really want to deal any further with your marketing efforts), the tests I agreed to do were the analog tests. The spdif signal checks I did were purely for my curiosity.
Tell you what, next time you ask the same question, I'll just say "5" and you can refer back to this post.
Tell you what, next time you ask the same question, I'll just say "5" and you can refer back to this post.
You seem to fail to understand or do so deliberately - the attenuators reduce reflections & therefore jitter (my premise). You were testing for jitter in the analogue domain yet you ensured that there were no reflections. Now what bit do you not understand? You invalidated the test, simple!
Argue with George- he said I *did* have reflections.
If by "ensuring there are no reflections" means using the correct (cheap, standard) coax and the correct termination, then I plead guilty. If that's all it takes to cause a $20 unit and a $500 unit to both perform equally well, then audiophiles can draw their own conclusions about the need for the $500 unit.
Argue with George- he said I *did* have reflections.
If by "ensuring there are no reflections" means using the correct (cheap, standard) coax and the correct termination, then I plead guilty. If that's all it takes to cause a $20 unit and a $500 unit to both perform equally well, then audiophiles can draw their own conclusions about the need for the $500 unit.
No, again your are trying to obfuscate (& trying to damage my commercial interests) - the attenuators are a completely separate & independent issue - they have nothing to do with the difference between the $500 & $30 device.
What I'm saying is that your tests on the attenuators are invalid. This was pointed out to you on a number of occasions & you refused to test their purported operation. What was the purpose of your test? Nobody ever denied that fixing up the termination of the input stage would be a better solution but not too many people have the desire to open up their DACs & the equipment to measure the impedance of their cable/DAC input stage.
This modus operanda casts a significant doubt over your following tests on the Hiface device.
Yes, I see in that test a lot of reflections, and SY had managed somehow to keep them for good.
When I have pointed out that with a minimum effort the signal should be much cleaner, then we concluded that it was not his primary goal.
Then we have arrived to this "general" statement that jitter in audio is just a bogus thing, a hype, for the religious and the "stealers". And anyway nothing is detectable, and the cheap chinese thing is just the same sounding than a better elaborated effort.
As told, I'm liberal - if this is your personal opinion, keep it.
In the meantime I'm back to fussing around with my micro -controlled VCXO driven dac, which refreshes the clock rate at every half a minute, so that the VCXO control voltage is practically a monstrously filtered DC level, - and still sounds different with these transports. And the ESS dac is no different at all in this respect.
And would like to find out about the "why" of that shitty sound of this beautifully measuring EMU soundcard. Which, by the way, sounded better in a contest when directly compared to a Behringer DCX.. (modified)
Ah, I had measured ~300psec rms jitter on it.. no, that can not be a problem, as we know..
On the other hand, I really has to agree with SY that consultancy / test works in a situation where commercial interests are lurking in the background - should not be taken for granted. So there has to be a limit somewhere.
Ciao, George
When I have pointed out that with a minimum effort the signal should be much cleaner, then we concluded that it was not his primary goal.
Then we have arrived to this "general" statement that jitter in audio is just a bogus thing, a hype, for the religious and the "stealers". And anyway nothing is detectable, and the cheap chinese thing is just the same sounding than a better elaborated effort.
As told, I'm liberal - if this is your personal opinion, keep it.
In the meantime I'm back to fussing around with my micro -controlled VCXO driven dac, which refreshes the clock rate at every half a minute, so that the VCXO control voltage is practically a monstrously filtered DC level, - and still sounds different with these transports. And the ESS dac is no different at all in this respect.
And would like to find out about the "why" of that shitty sound of this beautifully measuring EMU soundcard. Which, by the way, sounded better in a contest when directly compared to a Behringer DCX.. (modified)
Ah, I had measured ~300psec rms jitter on it.. no, that can not be a problem, as we know..
On the other hand, I really has to agree with SY that consultancy / test works in a situation where commercial interests are lurking in the background - should not be taken for granted. So there has to be a limit somewhere.
Ciao, George
Last edited:
- Status
- Not open for further replies.