Hi Dan
You realise that answering those kinds of questions is a red rag to some others here.
What the 2 x 3R3 plus 1uF does is open for discussion - but the 0.33F supercap I think is less of a mystery. Ted Smith, designer of PS Audio 'DirectStream DAC' says his power supply to the 'switch' of the delta-sigma modulator, points to a very simple fact, that there are only two states, 'on' and 'off' and it is the 'on' state that is of interest here. You are basically switching your power supply hard 'on' and any noise or aberration there he correctly says become jitter/noise and sounds like that. So he states that he filters (note, he does not say regulate) the supply down to 1 Hertz. Now he has a nice big box and we don't, so I thought about it and realised that a lot of stored energy resists change, like a flywheel does, and hence why not use a supercap? To give credit where it is due, I also found out that Coris was using supercaps there, but not sure what his reasoning was, perhaps different from mine. Just one final thought, and this is more controversial, but I know that the late Allen Wright, Jocko Homo and myself, are convinced that sub-one Hertz jitter is audible, in fact the worst kind. So it is hoped that those supercaps do better than 1 Hertz filtering. Dampening? Just a different way of describing the same thing?
About both together, what the post-DAC filter does, that extra clarity, from the bass up, improved imaging and stability of soundspace and tone, adding those supercaps seems to cement those qualities and make them even more obvious.
I am not sure if Ken has yet tried the supercaps. But it is interesting, considering how this thread has gone, that the supercaps in Posts #1 to #6 have gone unnoticed by the declamateurs here. Maybe they just never looked at that sudden burst of schematics that I don't think anybody has done on www.diyaudio.com before.
Maybe they might just hone in on them now?
Cheers, Joe
-
Last edited:
Does anybody know how to get an Audigy 2 ZS working as a 24/96 spectrum analyser? Tried various combinations of drivers and can't get past 48K.
I know that SY has one. Stuart, can you help me? I see that you have one of these cards, or did have. The ARTA test in chapter two shows reasonably low distortion - but need this to work up to 48K Nyquist.
Any help will be very appreciated - with Steve Bolser creating some files with stimuli, I want to try use this card and see what we can come up with. Steve has the Juli@ card and I am envious.
He is expressed a thought of doing the same with his card. But as promised earlier, whether successful or not, I will be posting. I promised full disclosure.
Last edited:
Well, I had a little bit different reasoning, but finally it is very near Joe`s. Providing actually an infinite energy capabilities to the circuit involved, it obviously improve the signal quality out of that device.
Switching mechanism in digital devices, and Joe`s explanation for using large decoupling caps on such it stands, and I agree, but there is a benefit of using the same approach for analogue circuits too.
I can confirm the above description of subjective improvements for the audio signal (sounds). Large caps in this case it even improve the already improved outputted sound when using this filtering cap over DAC`s differential outputs.
Switching mechanism in digital devices, and Joe`s explanation for using large decoupling caps on such it stands, and I agree, but there is a benefit of using the same approach for analogue circuits too.
I can confirm the above description of subjective improvements for the audio signal (sounds). Large caps in this case it even improve the already improved outputted sound when using this filtering cap over DAC`s differential outputs.
You don't get it, do you? It is not the "free" which concerns me, but the "it". I am unconvinced that there is anything to share. Let us be clear about this: I am not objecting to what you and others claim, I am disputing it. There is a world of difference. I am not saying "You can't say that"; I am saying "What you say is incorrect".Joe Rasmussen said:
My only complaint is that you and others refuse to engage with those who disagree with you. Instead of discussing circuits, you keep asking us to try it ourselves.
Let us suppose that I did what you ask, and heard a difference. What does that mean? Absolutely nothing. I would still have no further information to work with.
NO!!!!! How many more times do I need to say this? You do not need to personally experience an alleged effect in order to be able to provide a reasonable explanation of it. Two reasons I know this is true:Coris said:
1. circuit theory is about circuits, not human experiences of the output of circuits
2. many people have tried Joe's mod; as far as I know not one has come forward with a plausible explanation.
There seems to be two sets of people here: those who believe they have experienced the change, those who seek an explanation of the change. These two sets seem to be almost disjoint, not one and the same as you keep claiming.
A driver will not overcome the fundamental Shannon-Nyquist sampling theorem.
I used an Audigy about 10 years ago- better than nothing, but not very good. I would strongly recommend the M-Audio 192 as a good start, a Lynx L22 if you can afford it. And based on your question, you have some studying to do on basic principles before you even start, otherwise you won't be generating valid measurements. If you have AudioXpress, read my series on soundcard measurement, which is aimed at people who want to do measurement but aren't quite sure where to start.
If a proper test gave a negative result, then there is no discernable effect so no need to search for a cause.Ken Newton said:
If a proper test gave a positive result then there must be a change in the system output signal. The nature of this change may help us find the cause. However, I am doubtful of the value of a listening test in the absence of any clear ideas about the circuit changes themselves.
Careful circuit analysis could reduce the need for a listening test or alternatively be an input into designing the test. We can determine what circuit effect the circuit change would make, and then estimate what (if any) audible effect it might have.
@ pano & Anatech,
Ken is right on the mark with his questions and these do not imply any dishonesty.
Although i don´t know SY personally, i´d never question his integrity, which means that he would not manipulate an experiment.
But otoh he did express in this forum that he is a former "true believer" and an controlled listening experiment has been an epiphany after which he realizes that he did not hear any difference, so was transformed from "true believer" to "non believer".
With a scientific background you should be aware of a resulting bias instead of denying its existence.
See for example the impact:
All of a sudden it isn´t an experiment if a difference can be percepted but instead a test if a difference is as "dramatic" as it should be (at least for SY) to justify the usage of some words/phrases.
What about the power of the test? Nothing to worry about, because a "dramatic" difference must be heared in any case?
Further, it is obviously not a good idea to let a participant choose some details of a test protocol if that person is not experienced in doing controlled listening tests.
Without training under the specific condition such a test will most likely not work as it should...
Ken is right on the mark with his questions and these do not imply any dishonesty.
Although i don´t know SY personally, i´d never question his integrity, which means that he would not manipulate an experiment.
But otoh he did express in this forum that he is a former "true believer" and an controlled listening experiment has been an epiphany after which he realizes that he did not hear any difference, so was transformed from "true believer" to "non believer".
With a scientific background you should be aware of a resulting bias instead of denying its existence.
See for example the impact:
All of a sudden it isn´t an experiment if a difference can be percepted but instead a test if a difference is as "dramatic" as it should be (at least for SY) to justify the usage of some words/phrases.
What about the power of the test? Nothing to worry about, because a "dramatic" difference must be heared in any case?
Further, it is obviously not a good idea to let a participant choose some details of a test protocol if that person is not experienced in doing controlled listening tests.
Without training under the specific condition such a test will most likely not work as it should...
Last edited:
That should be the main reason of this tread... Your contributions to an explanation is only welcome...
The circuit theory it tells you how it works. Well, all right. But you have to agree that you do not know at all what is the differences in sound coming out of a standard device, and out of a device with this filtering on it. You just do not know that. That is a fact.
I, and few/many other know what that difference consist of. Another fact...
The same holds true for controlled listening tests. In the last 80-90 years a lot of scientific work was done wrt sensory tests. But, in a somewhat miraculous way most of it is omitted in listening tests in which controversial effects were under examination.
With that statement, I fully agree.
However, DF96 appears to be suggesting that we should simply procede to the search for a circuit based cause without first having determined that there is any subjective effect to find a physical cause for. I believe DF96 is only planning to perform his search effort on theoretical/simulation grounds, and not even perform his own measurements on the physical experiment. I assume this
because he continues to argue that personally performing the listening experiment is not relevant to the search for physical cause. Such views seem illogical to me.
What would be the point of searching for the physical cause an purported subjective effect without first having made some level of determination, anecdotally or scientifically, that the effect in question is subjectively audible? While such determination is ideally made via controlled listening testing using a large enough sample, I do not see that it is necessarily invalid for individuals to make such a listening determination anecdotally to their own personal satisfaction or dissatisfaction. Isn't personal anecdotal listening evaluation the DIY experimental norm?
You would do much better with ARTA, Virtins, or AudioTester. And a better soundcard. And an interface for the soundcard.
Most important is to understand the measurement you're making- modern computer-driven measurement will allow you to quickly and efficiently generate huge amounts of incorrect data.
Most important is to understand the measurement you're making- modern computer-driven measurement will allow you to quickly and efficiently generate huge amounts of incorrect data.
For the record I have an audigy which I bought cheap (as its 12 years old) to experiment with before dropping more on a card. the PC it was in then got repurposed by my darling offspring and I only just have it back. I would rip Vinyl and test my tube power amp with it. Would not use it for line level or DAC testing other than proving the UUT was better than the test set or the tester.
Agree, with the caveat of exactly what constitutes a proper test.
Also, agree.
Doesn't this point counter your own first point above, which was that if there is no discernible listening effect there is no need to search for a cause? Doesn't it instead make far more sense to first discern whether there is an audible effect, and only on a positive result then conduct a search for the physical mechanism causing that effect?
Hi Jakob2,
Many years ago I was measuring a circuit with a response that was too good to be true. Eventually I was able to trace the fault to a piece of measurement equipment that did not perform to it's own standards. I now know exactly what the problem was and can reproduce it even though I gave that piece of equipment away (that I bought new for north of $2,500 years before) and told my friend of that issue. The replacement is an HP 339A, and it agrees with expectations. The fact remains that I had faulty data and owned up to it, then traced the problem to the offending equipment and disclosed that.
Doubting Stuart I guess is Joe's right, but I think it is more fear based. Stuart could be Joe's greatest proponent if he would allow it.
One thing that greatly concerns me is the lack of suitable test equipment that Joe needs in order to do the testing of his own work and research. That he is modifying and assessing his own work without any form of proper testing just isn't right. I'm pretty shocked by this. I feel that Joe is not qualified to do this work without a basic test bench and a good signal analyzer. What he is doing places him in business doing work he can't check. This is work that by rights Joe shouldn't be doing. This goes for anyone in the same position. I'm not just picking on Joe here.
-Chris
No one said they did. But a listening test can guide the measuring direction as Stuart had said earlier. Like any branch of science, you follow the evidence no matter where it leads. Joe has doubts about this, and that is unfortunate.
Many years ago I was measuring a circuit with a response that was too good to be true. Eventually I was able to trace the fault to a piece of measurement equipment that did not perform to it's own standards. I now know exactly what the problem was and can reproduce it even though I gave that piece of equipment away (that I bought new for north of $2,500 years before) and told my friend of that issue. The replacement is an HP 339A, and it agrees with expectations. The fact remains that I had faulty data and owned up to it, then traced the problem to the offending equipment and disclosed that.
Doubting Stuart I guess is Joe's right, but I think it is more fear based. Stuart could be Joe's greatest proponent if he would allow it.
One thing that greatly concerns me is the lack of suitable test equipment that Joe needs in order to do the testing of his own work and research. That he is modifying and assessing his own work without any form of proper testing just isn't right. I'm pretty shocked by this. I feel that Joe is not qualified to do this work without a basic test bench and a good signal analyzer. What he is doing places him in business doing work he can't check. This is work that by rights Joe shouldn't be doing. This goes for anyone in the same position. I'm not just picking on Joe here.
-Chris
Hi Ken,
Changes to a product should also be measured to see how and if the deviations from normal functioning are greater than claimed. It behooves one to gather all available evidence before reporting the results. It makes no sense to stop short with any examination.
-Chris
No, not at all.
Changes to a product should also be measured to see how and if the deviations from normal functioning are greater than claimed. It behooves one to gather all available evidence before reporting the results. It makes no sense to stop short with any examination.
-Chris
I am the one out on a limp here, you realise that? Makes me the target that I never wanted to be - in the beginning did not realise it was going to come to that. I thought I would be taken at face value - and my ideas were open for anybody to test.
You realise that Ken was very skeptical in the beginning, but got curious - and found it was for real. So a little trust and a little less distrust is actually very healthy. I don't want to assume the Stuart lacks integrity, just as I would like to think he doesn't think I do either. But he needs to be a bit more convincing and I would feel a lot better. Accusing me of being interested in only money, that simply is not true. I just wish he would shelve that. But that's up to him, but yes, it effects my view of him, sadly.
Re SY and his Epiphany, Ken and the rest:
You know, we are often talking cross-purposes when we don't need to. It's about two different things, perhaps even three. They all have at least some validity, none are perfect.
1. Double-blind tests - some are believers, some are disbelievers, and some are agnostic. I am in the latter camp.
2. Personal anecdotal listening tests generally work over time. They allow for a more natural listening environment, they are not stilted nor regulated, some may even say strangulated
D), but for every wrong step there are often many right steps. Most advances in audio was achieved this way. Measurements generally comes later (there was a time we hardly knew how to measure speakers), then those led to further more focused development. Throughout the history of the civilised world, this has actually worked. 3. Measurements versus listening. To me this is a non-argument. We all want to explore, we all want to correlate, we all want to understand the science that lurks beneath. But most of all, we want to enjoy listening. Not all will ever come to a deeper understanding about circuit topologies and stuff like that - so to them it is also a non-argument. They just want to hear their system sound better. This is where people like me come in - and Stuart - nobody needs to get upset or fixated about what I/we do. Some very exalted names are on these forums who do the same and I am but a small cog in that scheme. Also a great way to make a lot of friends.
What I would like to see from this thread from now on is a more civilised discussion, less aggro, less dogmatism - not much to ask for.
Much of the heated argument is totally unnecessary.
- Status
- Not open for further replies.