So you think you set up an DBT and if you get a null result the "parameter may be too small to be heard" ?
? Did I say that ?
No, the parameter could not be heard by that listener in that setup. You can't extrapolate to the general case. This is all about the thing I say over and over again (and said to you several posts back, in #202)- you have to define the question that an experiment is supposed to answer before you can design a sensible experiment and correctly interpret the results.
Educated from a 10 page pamphlet maybe
The issues of wrong bits in audio have been resolved 10-20 years ago. Today's frontiers of digital deal with other effects and phenomena. Time to re-educate yourself perhaps?
Q&A with John Swenson. Part 1: What is Digital? | AudioStream
Q&A with John Swenson. Part 2: Are Bits Just Bits? | AudioStream
I wouldn't say USB audio is at the frontier of digital.
As to educating oneself on signal integrity, I would recommend people start with Howard Johnson, Eric Bogatin, Ralph Morrison etc true gurus of digital design and signal integrity.
I would also say that a lot that is referenced in those links is rather simplistic and geared very much towards audiophile beliefs.
Last edited:
It was too small to be heard in that test, or it may be completely irrelevant to sound quality. Further tests or theoretical analysis would help.gk7 said:
No. You gave some examples of claimed DBT results (subsequently disputed by others) which some audio fans might regard as being self-evidently daft, as a way to discredit DBTs. I was showing that what is self-evident to one person might be demonstrably false to another. You can't prove a point by stating something which itself needs to be proved. Your argument appeared to be:
DBT gave these daft results, therefore DBT is daft.
My response was: how do we know these results are daft? We can't simply assume this.
SY's response was: DBT did not give these results.
It's not a null result, and it is not unscientific.
If the DBT is performed on a subject that claims there to be a difference in sound between two cables and the result is that there is absolutely no correlation between these two (when using that person as the subject), the result is exactly "no correlation between cable and sound". That is not a null-result.
It also suggests that, assuming that the person is NOT lying, that the correlation lies elsewhere.
Some suggested correlations could be:
"money spent vs. sound"
or "what I've heard about this before vs. sound"
or "I like the look of this cable vs. sound"
Last edited:
I think I have.
Interestingly, if you weren't aware, in science, the opposite of what you are saying is "more true".
In other words, generally, if a correlation is found, it does not mean that it proves the cause. In other words, correlation != causation.
In other words, NOT finding a correlation whatsoever gives a much more trustworthy answer, because you don't have to speculate around what the real cause might be.
But this is a bit off-topic.
Instead of telling me to read up, please comment on the content/problems in my previous post instead.
Yes it "may". So what do we know after that test ? My point is that DBTs are a just one tool in the box,
not the alpha and omega of audio evaluation, completely in line with my post #225.
Yes, exactly.
Why "claimed" ? I posted the links.
I never said they are "daft". It is not helpful in a discussion if you interpret what I wrote.
So now, I did not say "these results are daft", to whom did you respond ?
Correct, correlation != causation. Now think through your own post again.
Wrong, a good example how not to make a deduction.
No sorry, I don´t see why I should this discuss with you.
gk7 said:
Perhaps I misunderstood what you were saying in post 225. I thought you had listed three DBT results which some on here would regard as self-evidently false. You did not use the word 'daft' but that appeared to be your intent. Perhaps you could explain, if I have misunderstood your argument?gk7 said:
I´m sorry that this drifts again into a pro/con DBT discussion.
I did _not_ dismiss DBTs as such in my original post.
I absolutely agree with SYs statement "subjective is not the same as irrational".
My point is that you don´t need a DBT to disprove an irrational claim (like
the original topic of this thread). It might even be not "the weapon of choice"
in fighting the increasing irrationalism of this "postmodern" age.
A nice evening to everybody, I have to leave now.
I did _not_ dismiss DBTs as such in my original post.
I absolutely agree with SYs statement "subjective is not the same as irrational".
My point is that you don´t need a DBT to disprove an irrational claim (like
the original topic of this thread). It might even be not "the weapon of choice"
in fighting the increasing irrationalism of this "postmodern" age.
A nice evening to everybody, I have to leave now.
Again you forgot to actually say what you had to say. Or do you have anything to say?
I was talking about the lack of correlation, it is of no correlation to the discussion that correlation != causation. That would be a logical fallacy.
It's not a deduction, it's a fact. No correlation means no correlation in the test that was setup.
"No correlation" does not mean "maybe a correlation", whereas "correlation" means "A is a potential cause to B" (if that's what is interesting).
All your "answers" are like that.
What exactly have you said, at all?
I wouldn't say USB audio is at the frontier of digital.
As to educating oneself on signal integrity, I would recommend people start with Howard Johnson, Eric Bogatin, Ralph Morrison etc true gurus of digital design and signal integrity.
I would also say that a lot that is referenced in those links is rather simplistic and geared very much towards audiophile beliefs.
Well it is true I think that those linked articles start with a premise that some of the timing related related effects John Swenson mentions cause audible effects. However Mr Swenson has from my observation on various fora over the years been a tireless experimenter with cause and effect; isolated and explored as much as possible (within the bounds of what an individual can do, i.e no blind listening panels). So whilst there remains an element of conjecture in what he says, he has in fact drawn on hard won experience and no small amount of expertise.
His articles seem to me to at least raise some interesting theories as to why bit correct transmission is not the only important factor in digital audio.
John Swenson's day job is power distribution in large chips. Given that, it is hardly surprising that he feels that this a neglected area of DAC design. Most of us emphasise the things we understand - especially the things we believe others don't understand. I think the most important point he makes is that slower logic can work better because it causes less ground bounce; some people seem to assume that faster is better.
I'm not sure what this adds to talk about USB cables.
I'm not sure what this adds to talk about USB cables.
I would have thought that ground wire resistance and inductance would be a factor when considering ground bounce.
Look up the people I mentioned, and add Henry Ott.
What he sais is no great revelation to those that work in electronics (and if it is then maybe its time to move on). I have been laying PCBs out for over 27 years, and in all that time delta I noise on ground (and power) planes has always been a concern, and with todays high speed interfaces and the demands of thousands of gates switching it is even more critical to get the design right. It is called SIGNAL INTEGRITY, it is not new, it is part of the requirements of engineering a particular circuit from conception to working product.
I have as well as the generic layout software a signal integrity verification tool and a power integrity tool, that allows us to simulate the layout and get the signal integrity right.
Bit transmission is the most important thing in digital, but there is a lot of hidden engineering round getting this bit transmission correct, it is call signal integrity.
Here is one of the main sites for this info:
Signal Consulting, Inc. - Dr. Howard Johnson
On a second note, I often mention things that I do in the real world regarding signal integrity, power supply integrity and EMC, but it is often ignored or decried because it doesn't often fit with the current audiophile design fashions, yet it is good enough for very high reliability products.
So while his articles are OK it is not new, and instead of chasing Maxwell's demons, apply good digital engineering practice and these things can be controlled. Same with analogue/digital designs and RF/digital designs, there is a wealth of information out there.
Finally for digital (and analogue) you don't use wires you use planes on PCBs and it is inductance that is the most important factor.
What he sais is no great revelation to those that work in electronics (and if it is then maybe its time to move on). I have been laying PCBs out for over 27 years, and in all that time delta I noise on ground (and power) planes has always been a concern, and with todays high speed interfaces and the demands of thousands of gates switching it is even more critical to get the design right. It is called SIGNAL INTEGRITY, it is not new, it is part of the requirements of engineering a particular circuit from conception to working product.
I have as well as the generic layout software a signal integrity verification tool and a power integrity tool, that allows us to simulate the layout and get the signal integrity right.
Bit transmission is the most important thing in digital, but there is a lot of hidden engineering round getting this bit transmission correct, it is call signal integrity.
Here is one of the main sites for this info:
Signal Consulting, Inc. - Dr. Howard Johnson
On a second note, I often mention things that I do in the real world regarding signal integrity, power supply integrity and EMC, but it is often ignored or decried because it doesn't often fit with the current audiophile design fashions, yet it is good enough for very high reliability products.
So while his articles are OK it is not new, and instead of chasing Maxwell's demons, apply good digital engineering practice and these things can be controlled. Same with analogue/digital designs and RF/digital designs, there is a wealth of information out there.
Finally for digital (and analogue) you don't use wires you use planes on PCBs and it is inductance that is the most important factor.
USB signal transmission it is diff pair, which makes it pretty immune from ground bounce and have better timing.
http://www.ieee.li/pdf/essay/differential_signals.pdf
http://www.ieee.li/pdf/essay/differential_signals.pdf
Bit transmission I am sure is the most important thing in digital. But for the life of me I can't see why this is even relevant to this discussion. In digital audio, jitter is important too. This thread is about the possible audibility of changing a USB cable. I would infer that in many cases (nearly all if you credit Mr Swenson's actual investigation of audio DACs) the USB cable links to a device which has some deficiencies. In those circumstances it is not beyond the bounds of possibilty that two USB cable with different characteristics will sound different, for the reasons explained by Mr Swenson. In many cases the power grounds are linked between computer and DAC by the USB cable.
Last edited:
It's the only thing. When talking about cables.
The discussion is whether two USB cables produce the same audio data. Which they do. And that has to do with whether or not bit transmission is flawless. Which it is.
In this case, it definitely is not.
Whoever clocks the data doesn't care about the cable. Thus, the cable has nothing to do with it.
Yes?
Are you insane?
Yes, it is beyond the bounds of possibility. Especially when talking about Hifi-equipment specifically designed with noise-reduction in mind, rather than intentionally using attached cables as antennas to the best of their abilities.
But this doesn't even have anything to do with jitter, which seemed to be your concern?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.