| Columns Retired Columns & Blogs |
| Loudspeakers Amplification Digital Sources | |
Analog Sources Accessories Featured | |
Music Columns Retired Columns | |
Show Reports | |
Features Latest News Community | |
Resources Subscriptions |
© 2024 StereophileAVTech Media Americas Inc., USA
All rights reserved
Perhaps you have something constructive to add, rather than trying to add more insult to the people who have the right of this discussion, Mr. "frog". Again, it's not surprising that you hide behind a pseudonym.
As I said, I see no reason to doubt my original analysis of the quote originally posted.
Those of you who still think that one can not "resolve" where, for instance, something starts, to well under a sample would be well advised to simply go try it.
I see once again you don't have anything constructive to add, and would rather try to add more insult to the people who have the right of this discussion, Mr. "j_j". How fitting, that you have decided to end this thread with yet another unprovoked personal attack, instead of showing some class, or at least pretending to sound like a reasonable person for a change, and ending your hostilities on a high note. Your dishonesty and desire to mislead readers throughout this thread has never ceased for a minute, and continues even after this debate is now over, with a lying fabrication about my name. Not only do you provide no proof for these hit and run defamations of yours, ever, but likewise, you provide no proof when asked to in your debates, as well. People here are getting all too familiar with your misleading games and deliberate dishonesty, "j_j". The fact that you would even try to defame another member with a lie about "hiding behind a pseudonym", while you refuse to even say who you are and are clearly hiding behind a pseudonym, means that we can pretty much dismiss anything you have to say here as irrational nonsense, from a "senile old curmudgeon" (as you have described yourself in the past).
Of course you don't. As you are an infamous dogmatist who has never been known to debate honestly or objectively, no one was ever expecting you to change your position, after having wasted all our time on these useless "analyses" of yours, which didn't even follow what was being explained in the papers you debated (but like a true "pseudoscientist", never even read!). Unfortunately, now that the author of the papers himself has set the record straight, everyone has reason to doubt anything you have to say on the issue. Ergo, you have nothing further to say on the issue. Live with it, deal with it, accept it. Move on, and move out. Or if you prefer, Mr. "J_J", stick a fork in yourself, 'cos you're done here!
n.b. Until you reveal your full and real name on this forum, do not expect to be taken as anything but another faceless, nameless, agenda-driven troll from the Hydrogen Audio Cult forum. It's a good thing you're clearly not a real scientist, as you like to pretend to people that you are (which is like Guy Caballero pretending to be handicapped...), because I think you've turned in a truly shameful performance and behaviour in this debate; which no real scientist would want to stick his name to. No kidding you refuse to reveal your real name here!
Have a nice life, j_j.
Sorry, I've been bizzy the last couple days, and oddly enough, flamewars are making me go zzzzzzzzzzzzzzzzzzzzzzzz.
So, obviously, if you've got any filter F (or any LTI system for that matter) and an input X, and output Y=F[X], obviously one could define the "magic" M for which X=Y+M, that is, M=X-F[X]. And saying that X "contains" M for any arbitrary filter is obviously false in any meaningful sense, as you are supposing - clearly if I define, say, an equalizer filter or an integrator or whatnot, it doesn't make any sense to say that the output is "contained" in the input.
But we're talking about a very specific LTI system here - specifically the antialias filter before the sampling operation. Ideally this filter does nothing more than "accept anything under Fs/2 at gain=0db and reject anything at Fs/2 or above with attenuation=infinity". That is, in the frequency domain, all the amplitude/phase components at f<Fs/2 are unmodified (and zero above that).
I assert that, if having one signal "contain" another signal ever really means anything, this situation would fit - the continuous signal X really "contains" the antialiased signal Y, and by extension, the discrete time signal Y[T]. Just like a piece of candy can be separated by the wrapper containing it (or if you want a negative connotation, like a garbage bag contains garbage), given a sampling rate, one can clearly and uniquely separate the frequency components under half that rate (Y) from the "wrapper" or components over half that rate (M), and you can combine those two to get the original signal (X), which contains Y. You don't need some capricious or arbitrary filter to go from one to the other. It's not magical.
This all sounds like crazy wordplay now that I read it back over in my head, but it's really key in the way I think about digital audio. What is PCM-sampled really is contained in the original signal. While it is not a perfect representation of it, up to the limits of sampling rate and quantization error, the frequency components are exactly the same. And I interpret many people as saying that those limits are far more dangerous and fundamental than they really are. So, when I see a statement to the effect of "16/44 cannot encode two analog pulses 5us apart" or "16/44 cannot encode a 5us pulse" or whatnot - well, that's wrong, because it really can encode the pulses - albeit only the bits under 22khz!
And that belief transforms the debate into something to the effect of, if we're so concerned about the representation of these pulses at 16/44, but what we're really wanting are the frequency components above 22khz, how important is that? How does that square with the tested (in)audibility of ultrasonics? It's not a question about whether or not this or that can be represented.
Well, by "pulling his leg" I really meant to say "trying to flame the fuck out of him using whatever flimsy basis they can comprehend". As opposed to, you know, actually not having enough of a grasp of the internet (or this website, for that matter) to figure out the real names of each of us.
I'll cop up to having an extremely naive nature, actually. And as my late posts imply I've been kinda out of it recently.
I can use that justification for everybody else's behavior, but not my own.
Besides, it just doesn't seem fun to me to simply call somebody a 'braying ass'. That's just too blas
Except that none of those authorities, with the exception of certain engineers, could be *expected* to be expert in digital sampling theory as it applies to audio recording and playback.
In some professional circles, so far. It would be interesting to see Dr. Kunchur's paper (is there one, btw?)
on this matter reviewed for JAES publication, for example.
Also, some of us *are* scientists, (though not necessarily in fields related to audio), and as such are familiar with the process by which research becomes publication -- including its strengths and weaknesses. That might be good for Dr. Kunchur to know, so he doesn't have to assume we are all naive about the process. Some of us are also well aware of the pitfalls of getting one's information from nonauthoritative online sources.
It was good to see Dr. Kunchur acknowledge the limit of home audio 'testing' ( though his implication of IRBs as a validator of claims is a little strange. And IRB panel would not likely catch a mistaken idea about digital sampling, for example; its focus is on protecting the rights of experimental subjects) I'm also very happy to see Dr. Kunchur warn against reliance on the 'intuitive' or 'common sense' in digital audio -- a common mistake that audiophiles make -- because some central ideas of digital audio are actually non-intuitive (like the Shannon-Nyquist theorem).
I'm sorry if this gets taken to private mails...I would like to see the apparent conflict between what those on this thread who have proven themselves to be knowledgeable, professionally, about sampling theory and human hearing, and Dr. Kunchur, are resolved.
(I'd also bet MJ Frog understood nary a word of the technical information the Dr. Kunchur provided, and for which he was so thankful.)
EDIT: I see ncdrawl has sensibly started a new thread to address questions to Dr. Kunchur.
Axon, I suggest you read and study Dr. Kunkur's papers again.
-----
First, let's take a look at AES and its backing. Maybe take a look at who is not. I don't see Sony's name.
Notice the names of audio companies who back AES.
Sustaining Member Organizations:
http://www.aes.org/sustaining_members/
"A&R Cambridge Limited
ACO Pacific, Inc.
Acustica Beyma S.A.
Air Studios Ltd.
AKG Acoustics GmbH, a Harman International Company
AKM Semiconductor, Inc.
Amber Technology Limited
Anchor Audio, Inc.
Apogee Electronics Corporation
ATC Loudspeaker Technology Ltd.
Audio Limited
Audio Media/IMAS Publishing Ltd.
Audio Partnership PLC
Audio Precision, Inc.
Audio Pro International
Audiomatica S.r.l.
AudioScience, Inc.
Autograph Sound Recording Ltd.
B
B & W Group Ltd.
Banff Centre
British Broadcasting Corporation
C
Cadac Electronics Plc
Calrec Audio Ltd.
Cambridge Silicon Radio
Why? JAES would seem to have their own biases. How many articles has the JAES published on the effects of cables as described in the audiophile press?
I might have missed it but I don't see any audiophile cable manufacturers listed as supporters of JAES.
Would this be just a simple coincidence?
Except that none of those authorities, with the exception of certain engineers, could be *expected* to be expert in digital sampling theory as it applies to audio recording and playback.
You again. Back again here, with your usual assertions without proof. Which is all we have seen from you people. According to way you pathological skeptics conduct yourselves, "expectation bias" is only something the rest of the world needs to be concerned about.
In some professional circles, so far. It would be interesting to see Dr. Kunchur's paper (is there one, btw?)
on this matter reviewed for JAES publication, for example.
I think it's quite telling that after this debate is over, you saunter back in with your usual personal attacks, to ask if Dr. Kunchur has even written a paper. Thanks "Krabapple", that sums up your "contribution" to this debate quite nicely.
Also, some of us *are* scientists, (though not necessarily in fields related to audio), and as such are familiar with the process by which research becomes publication -- including its strengths and weaknesses.
Fully none of you are scientists. That has been all too evident. Do not confuse "pseudoscientists" with "scientists". And here's another newsflash: everyone in this debate is familiar with the process by which research becomes publication. So what? It has nothing to do with all the wrongful assertions you dogmatists have made against Dr. Kunchur's paper.
I'm also very happy to see Dr. Kunchur warn against reliance on the 'intuitive' or 'common sense' in digital audio -- a common mistake that audiophiles make -- because some central ideas of digital audio are actually non-intuitive (like the Shannon-Nyquist theorem).
LOL! Obviously, you didn't understand enough of what was said here, to realize Dr. Kunchur was arguing against the Nyquist-Shannon theorem.
I would like to see the apparent conflict between what those on this thread who have proven themselves to be knowledgeable, professionally, about sampling theory and human hearing, and Dr. Kunchur, are resolved.
Rest easy. It has been resolved. By the fact that there was no "apparent conflict" to begin with; since no one in this thread arguing against Kunchur's studies has proven themselves to be "knowledgeable, professionally, about sampling theory and human hearing". There's no reason for you to continue being here, Krab.
(I'd also bet MJ Frog understood nary a word of the technical information the Dr. Kunchur provided, and for which he was so thankful.)
And BANG! The predictable personal attacks from the Krabapple troll of Hydrogen Audio Forum, come without provocation, and without restraint, and continue despite the fact that the debate here is over, and even the Zamboni operator has left the building. It is once again, telling that you are making another assertion without proof, because this is the behaviour that you, j-j, and the rest of the hostile debaters from Hydrogen Audio have consistently demonstrated throughout this "debate" (and I hesitate to call any discussion you are engaged in a "debate"). It shows why nothing that any of you have stated here is worth a plug nickel, how it is all about you guys stroking your egos and pursuing your anti-audio agendas, and your biased assertions also show why you don't know the first thing about science.
Unlike you, I know you didn't understand a plumb word of the technical information provided by Dr. Kunchur. And I'm not just talking about the fact that you demonstrated numerous times how you know nothing about audio, and unlike some of your audio cult brethren, you don't even pretend to have a scientific background in audio. I'm talking about the fact that unlike you, I actually read Dr. Kunchur's papers. Whereas as you just showed, you thought there was only one, and you weren't even sure he had written a paper, and you didn't even get what he was saying about the Nyquist-Shannon sampling theorem! I know you like to pretend you are knowledgable of subjects you have zero knowledge of, hence the reason you make these silly appeals to authority, pretending you have credentials you don't. But see, if you were among real scientists my dear Krabby troll, you wouldn't even get far enough with these proof-by-assertions of yours to be serving coffee to them.
If this garbage is all you have to "contribute" Krabapple, stay where you were and stop polluting our forum.
Thanks for the explanation, Axon.
If I read that correctly, all that says there is nothing above what is allowed by the sampling rate which is further assured by the antialiasing filter (which cannot be perfect or brick wall). Given what is there under 22kHz (in this case), you can mathematically calculate what should be above the filter - but it is not there, never will be and never could be. So, yes, this would appear to be some bit of word play to say the original analog signal is "contained" within the 16/44 output. It's estimating the number of beans in a jar.
What you've described sort of sounds like, "If you show me a picture of your entire family, I can tell you how many children are in your family even if they are not present because the representaion of your entire family is in the photo you're showing me and I can count the number of heads in the photo."
Which to me means 16/44 sampling is not providing sufficient resolution of any information above 22kHz when compared to higher resolution sampling rates. This would be possible only under ideal or theoretical and not real world conditions if all of this operates like most things do.
Going back to sas' graph this would seem to give an edge in frequency response and therefore time resolution to any format higher than 16/44 and definitely points to DSD being the highest resolution whether you can count how many kids are in the photo or not. This would almost certainly be true in real life applications.
How can counting heads be good for audio quality? That would seem to make me work much harder to put together what I'm hearing and that would be a main complaint with 16/44 that has been addressed with DSD and higher sampling rates.
So all of this has still been about what is "just barely good enough" under ideal conditions?
none.
P.S. I've only read the last page of this thread so don't get on me for not getting on JJ. Don't know what he's done....
Well, when you put two pulses separated by a time delay, you create a comb filter, and cause frequency shaping that is certainly measurable, and perhaps perceptable, under 20kHz. So the "two pulse can not be distinguished" claim is trivially shown wrong.
The argument about time vs. frequency resolution ignores basic understanding of tradeoffs between the time domain and phase in the Fourier domain.
There is little, if any, more to say about the response. There are still what looks like obvious mistakes, despite what Michelle Jay Frogger, or whoever the sock puppet is, has to say.
I have had a couple of discussions via phone with the good Dr. and am consolidating answers from here and a couple other forums to present to him. He will let the questions accumulate and respond in due time.
For simplicity's sake, to avoid his being bombarded, please present all questions to the thread that Ive started...no sense in his being bombarded by too many people. I am acting as the middleman for correspondence(with his permission, of course)
http://forum.stereophile.com/forum/showflat.php?Cat=0&Number=69464&an=0&page=0#Post69464
Wrong, JAES published an article about cable effects by Dr. Greiner.
Interesting approach, Frog. What are your qualifications for determining who is a real scientist and who is not?
How about posting your C.V.?
Hint: Having the same initials as a Stereophile columnist doesn't confer that sort of authority. (or do we even know what Mikey's middle initical is?) ;-)
JJ, don't waste your breath. It's quote clear that when we start talking about science and facts we're shooting right over many of these people's heads.
Experimental evidence means nothing to them because they apparently can tell whether it is valid or relevant or not.
Unless we start writing like "Wow dude, I put my speaker cables on telephone insulators and it took 13 veils off the soundstage..." we're wasting our breath.
Citation please. Which volume? Title of article?
Back at you. Greiner, cables, jaes gives way to many hits for me to bother copying a URL. You might even find a pdf of the paper.
Back at me? You made the reference. Have you actually read the paper you are citing? Until you can make a citation it would appear that "none" was the correct answer.
JJ, don't waste your breath. It's quote clear that when we start talking about pseudoscience and pseudo facts we're shooting right over many of these people's heads.
Yes, it's "quote clear" Kroogie. It certainly is quote clear.
Experimental evidence means nothing to them because they apparently can tell whether it is valid or relevant or not.
I don't know who you're insulting now, but that makes no sense. As ususal. Keep up the good work. Have some more bennies for breakfast while you're at it.
Unless we start writing like "Wow dude, I put my speaker cables on telephone insulators and it took 13 veils off the soundstage..." we're wasting our breath.
Well said. You're not just wasting your breath, you're wasting your time. You obviously won't find any converts stupid enough to join your pseudoscience audio cult on this forum. Which begs the question, Kruggie:
Why are you and the sockpuppet "j_j" wasting your time and breath here?
Which itself begs another question: I noticed that Dr. M.N. Kunchur, what we call in the business a "real scientist", was too busy to involve himself in this forum. So he asked that questions be emailed him, and he might get to them when he had the time. This is to be expected of any productive scientist. So how do both of you lying goofs, who troll your pseudoscientific propaganda on audio forums pretending to be scientists (in the hopes of obtaining credibility on spouting your pseudoscientific rubbish), explain the fact that you have all this time and breath to waste arguing with audiophiles you hate on forums across the internet, where the two of you are always regarded as (and given no more respect than) flies on shit?
This proves what I have said many times here. The well known trolls from Hydrogen Audio, "Krueger" aka "Kroogur", aka "Arnold" and his sidekick "j_j", aka "Woodinville", aka "Carl Curmudgeon", always engage in these pseudoscientific sleight of hand tactics. Whereby they demand proof from their critics, but when asked to provide proof for their fallacious assertions, they either ignore your request for evidence, or it's always responses like "back at ya". "I'm too busy pretending I'm a scientist on the internet!". That sort of thing.
So they can make any unfounded claim they want, but you see, the "burden of proof" will always be on you, even for their claims!!
Now we understand the reason why Krueger and j_j are regarded as little more than a pair of rubbish-spouting clowns on internet audio forums, who provide nothing good to a discussion group but hostile debates and flame wars, as they troll from one end of the net to the other on their anti-audio propaganda crusade.
You obviously have no concept of how to do academic research, or any kind of research for that part. The Greiner paper about cables is a classic, one that you obviously knew nothing about.
Of course your first mistake was making a negative hypothesis. Your second was making a poorly researched claim about the AES in front of people, some of whom obviously know better.
I make the reference and if it is incomplete, you at least google it so you don't look even more poorly-informed. Get it?
If you can actually log onto google and check it out, you will see that I may have read it before you were born.
I certainly read it before you learned to read, seeing as that appears to be an emerging capability for you.
You're obviously talking out the back of your neck. Do your homework - maybe a little education will teach you something about honoring your elders.
Oh, I get it, you don't have a library card so you can't look the paper up in the AES annals. Well, that explains the lip, as well as the ignorance.
Pure ad hominem and zero substance. So you got nothing. No surprise there.
*That's* your idea of logic? AK and JJ won't hand-lead you directly to a reference by Greiner about cable in the JAES, so they 'got nothing'? You really think that follows?
You're wrong, but you're still out of luck as regards having the reference handed to you. Because now it looks like you aren't really interested in finding it. You're going to have to show a little effort.
So here's the JAES search page:
http://www.aes.org/journal/search.cfm
If you can't figure out how to quickly find an article by 'Greiner' that discusses 'cable', let me know. It's a classic btw. (Hint: check the 'Search abstract' box).
But if you aren't willing to put at least this tiny amount of effort into your research, it would seem you aren't serious.
Actually, I have the PDF for the Greiner paper on my hard drive, where it has been for years along with every other AES paper through the early part of this millenium.
Proof? When and if you ever read the paper, you'll find that the last sentence is:
"The use of "special" cables, including normal coaxial
cable, is not warranted except in a few extraordinary appli-
cations. And in those applications, short runs of cable
would be a better solution."
It follows from posturing and intellectual laziness.
This is one reason why I say that many of the GE's who post here aren't worth the trouble to deal with.
Instead of facts and logic they have redicule and arrogance.
They want us out of here so that the discussion will go back down to their mental level.
It's the same tactic that the Nazis used in Germany during the 1920's. Chase out the intelligent people so that they can look smart.
Let's take a look at the record Arny. Unfortunately, you, J_J, Krabapple, Axon got caught red handed misleading the public several times. Here are a few.
From Steve, responding to dissidents who attempted to discredit Dr. Kunkur by claiming Dr. did not know basic sampling:
Scott wrote:That's* your idea of logic? AK and JJ won't hand-lead you directly to a reference by Greiner about cable in the JAES, so they 'got nothing'? You really think that follows?
It follows from posturing and intellectual laziness.
Which is exactly what you and your fanboy, the anonymous troll "krabapple" have demonstrated, by trying to shift the burden of proof on to your opponent. So once again, you make unscientific statements that have no evidence, and knowing this, you deliberately try to mislead readers into thinking there is evidence. This is another dirty deceitful debate trick I've seen you use a thousand times. This time, the attempt to mislead is to make the reader believe there is evidence to support some claim you made up by trying to hint at it in a credible-sounding fashion, then saying "Google It!". Funny how you and the ignorant blowhard "krabapple" both take the time to blast your opponent for pointing out that you are unscientifically shifting the burden of proof, but despite all your efforts to attack your opponents for not finding it for you, neither of you two goofs can come up with that paper.
Scott wrote:...if you aren't willing to put at least this tiny amount of effort into your research, it would seem you aren't serious.
This is one reason why I say that many of the GE's who post here aren't worth the trouble to deal with.
And you're right. Because even with this major slap in your collective faces, and your profound ignorance of audio science exposed by a genuine working phd'd scientist, the trouble has only begun for you Hydrogen anti-audio cult trolls.
So why is it you keep ignoring this basic question I have asked you: why are you Hydrogen Audio Cult trolls still here? Freddy, if you still think you're going to sucker converts on Stereophile into joining your pseudoscientific audio religion on Hydrogen, after this major fiasco where you, "j_j", "Krabapple", "Axon", et al. were shown to have attempted to mislead the public with dozens of erroneous statements on digital sampling theory, you're even stupider than I could have imagined.
Instead of facts and logic they have redicule and arrogance.
Facts and logic were well in evidence when Dr. Kunchur submitted facts and logic on his paper, thereby handing your collective asses over to you on a plate. I believe the well deserved ridicule you guys received came afterward.
They want us out of here so that the discussion will go back down to their mental level.
Arny, you make Ethan look smart and well-informed. We want you Hydrogen forum-bashing, propagnada-dispensing trolls out of here in order to have some chance of having audio discussions in peace, without the excessive hostilities and ad hominem attacks, which is all that you, krabapple, j_j and crew are good for.
It's the same tactic that the Nazis used in Germany during the 1920's. Chase out the intelligent people so that they can look smart.
I was wondering when you were going to get to the Hitler thing again. Given your well known family background in Germany, I would think you would have the sense to avoid comparing Stereophiler's with Nazis. But I see that's all you Hydrogen Audio boys have left, in your little bag of tricks you use to try to fool people. Thus, I invoke Godwin's law upon you. Thus ends the debate in this thread, whereby you and your ignorant wanna-be scientist cult buddies have now lost it in GRAND fashion.
Reductio_ad_Hitlerum: http://en.wikipedia.org/wiki/Reductio_ad_Hitlerum
LOL!
I'd like to see you document that "...well known family background in Germany", Frog.
My particular offshoot of the Krueger family hasn't lived in Germany since the early 1700s.
Got any more fantasies to share?
All the effort you and Arny have made to posture over this could have been reduced to a simple answer to the original question. Why all this effort not to answer the simple questions about this paper? Is that your idea of being "helpful?" By the way, Before ever asking for the title of the paper I had already googled "Greiner" and "cables" and found nothing about this paper. I had already gone to the AES search engine and found nothing under "Greiner" as an author. So what do we have here with your post Steve? Just more "bullshit" ad hominem with zero substance.
Thank you for that detailed summation of this debate, SAS.
It is very educational I find, in showing the true nature of the trolls from Hydrogen Audio who have invaded our forum (Arny, "j_j", "krabapple", "Axon", etc). Unlike other members of this forum, they all have a collective purpose in being here, behave like trolls and share the same common mindset. Both dishonest and hostile in their debates, their raison d'etre is to deceive the reading public with just enough smoke and mirrors to buy into their propaganda. Much as Scientologists will set up a store front to persuade the casual public to entire into their world of pseudoscience and religion. Hence the reason I often call the Hydrogen trolls "DBTologists".
With a large collection of "dishonest debating tricks", they ply their debating trade, in an attempt to get people to join their religious crusade against high end audio, backed by mediocre-end audio companies, which support the AES. The "smoke and mirrors' might be in the form of evidence that they pretend exists, or technical arguments that to some lay people might "sound good enough", but which as we have seen here after Dr. Kunchur exposed their ignorance, are only arguments that prove to be pseudoscientific rubbish. It's all about having to remove the facade that these Hydrogen trolls keep putting up. Once you expose them, you expose their agenda.
And now that their agenda has been thoroughly exposed in this thread, now that their pseudoscience has been revealed, now that their dogmatism has been demonstrated, now that their ignorance has been proven, now that their unfounded arrogance, ad hominem attacks and hostile debate tactics have been clearly laid out in this thread, I don't think things will be the same as they were before.
Every member on Stereophile who plowed through this thread can now see the intellectual bankruptcy, amoral behaviour and pseudoscientific dog tricks practiced by the audio cult from the Hydrogen forum, to deliberately and methodically spread disinformation among the audiophile community. Any last vestige of trust they might have hoped to create with readers was shot to pieces by Dr. Kunchur's response. Any credibility "jj", "Krueger", "krabapple", "axon" et al. might have hoped to create for themselves as "scientists", was also shot to hell when they lowered their reputations even further by attempting to debunk and discredit the years long work of a real phd'd scientist, with nothing more than bubble-gum doctrine. This teaches us that fake, pseudoscientific wanna-be blowhards on the internet, barging into audio forums and crapping all over the place, can in no way shape or form be mistaken for real audio scientists.
The only hope now for the Hydrogen Audio Cult of Pseudoscience is to go after new future members of this forum. Which I predict will be nigh on impossible to do, after they have been exposed by real scientists as pseudoscientists, and their reputations have been shot to hell, and their reputations for dishonesty and misleading the public have been well proven and documented here. So now that they have outstayed their welcome, I predict their only purpose remaining here will be to create unproductive flame wars.
Let's take a look at the record Arny. Unfortunately, you, J_J, Krabapple, Axon got caught red handed misleading the public several times. Here are a few.
From Steve, responding to dissidents who attempted to discredit Dr. Kunkur by claiming Dr. did not know basic sampling:
As one can see, all of these guys made huge errors, that Dr. Kunkur responded to.
Here is another one from J_J.
Notice no time delay figure is given. Now watch his broad, sweeping conclusion....
So the "two pulse can not be distinguished" claim is trivially shown wrong.
1)Who made the quote "two pulse can not be distinguished"
under 20khz?? Misleading again J_J. No wonder J_J doesn't sign his legal name to his own posts.
2) J_J makes a huge, sweeping, overall claim
with little specifics given to the test. What real scientist leaves out critical information so we cannot understand how he came to his conclusion. And notice, no proof yet again.
It is called, leave out critical information tatic, so one can mislead the public to a misleading conclusion.
Do you really want to trust people that use an assortment of tricks in an attempt to dupe you.
Krabapple:
Of course no explanation, just a charge to inspire doubt of Dr. Kunkur's work. Afterall, one has to have dozens if not hundreds of engineers etc agree before publication. See my next post with Dr. Kunkur's explanation of the rigors of the research/how his paper came to be published.
Steve: Better check Dr. Kunkur's correction of Arny Krueger's comment on page 1 and J_J's on page two. Mainstream science backs up Dr. Kunkur's papers.
Dr. Kunkur states (post #69424 on page 30 of this string:) Referring to Arny Krueger and J_J.
Let's go on.
Who was being corrected Krabapple???? Can you find where Dr. Kunkur quoted from??????
Luckily Dr. Kunkur is main stream, third party research.
As one can see, all three of these guys made huge errors, whom Dr. Kunkur's corrected.
Here is another one from J_J.
Notice no time delay figure, just "certainly measurable, and perhaps perceptable, under 20kHz." is given. But next he makes a huge sweeping conclusion....
"So the "two pulse can not be distinguished" claim is trivially shown wrong."
1)Who made the quote/claim "two pulse can not be distinguished" under 20khz???? Please show us who stated this J_J.
No wonder J_J doesn't sign his legal name to his posts. Worried about ethical violations as a minimum?.
2) J_J makes a huge, sweeping, overall claim
with little evidence. Is he still talking two channels, as he has for dozens of pages, or has he suddenly changed to one channel. One time he believes this, and another time he believes that.
He leaves out critical information and then makes a huge sweeping conclusion that misleads the public.
Do you really want to trust people that use tricks in an attempt to dupe you.
I see, so you would rather resort to pure ad hominem than actually answer simple questions that would take little effort. If you want to Email me a copy of the paper, I'll be happy to read it. S888wheel@aol.com
Well actually I thought I said the other way around - that the 16/44 output is "contained" in the analog signal. Obviously extrapolating what is above 22khz from what is below 22khz is guesswork in the general case. But when we're talking about the specific case here - about pulses - that extrapolation happens to be really exact.
It's a sort of analogous situation to the deconvolution problem in astronomy, where the effective resolution of the imagery can be dramatically improved given knowledge of the point spread function. ie, given a star that ostensibly is supposed to be a point light source is imaged as being much larger, the image can be deconvolved to make the star point-like again (and the rest of the imagery similarly improved in resolution) The PSF is equivalent to an impulse response in signal processing. And for digital audio, it's analogous to the ideal sinc function which performs the band-limited interpolation between samples at the analog output. So, in principle, if you have pulses spaced 5us apart represented at 44khz, you could upsample the signal, deconvolve it with a sinc filter, and get your original pulses back spaced 5us apart.
Well, sure. Like I said, in the general case (not making any assumptions about what signals we're talking about), there's nothing under 22khz that can tell us what's over 22khz. It doesn't have anything to do with the non-ideality of the implementation.
But that's not what we're talking about. For pulses, we really do have knowledge of what's there above 22khz, so we can identify the original pulse locations anyway. For my comment about the 16/44 signal being "contained" in the analog signal, it's all a matter of looking at the signals in the frequency domain rather than the time domain, and noting that the 16/44 frequency info is simply a subset of the analog frequency info. Again, wordplay, but to me it implies that the properties maligned here about 16/44 - notably the wide impulse response in the Pyramix plot - are categorically not distortion. They are what they are, and the specific technical reasons why they differ from the analog pulse (frequency response; transition band width; quantization error) are what should be being debated. Not the shape. Because that's more or less distortionless, and there is no reason to believe it is actually problematic.
More generally... what I'm saying is, give me any valid test that numerically estimates the "time resolution" of an audio format, based on the estimated timing of signal events, and I will show that 16/44 will yield a time resolution estimate that is way, way below 5us.
Now, reading a little more deeply into Kunkur's papers, he's actually arguing a significantly different point (besides all the mistakes I think are being made with regards to the properties of digital audio). He is proposing that nonlinearities in the ear are what causes ultrasonic differences to become audible. That's somewhat novel insofar as I'm aware of speaker nonlinearities to be proposed before as an audibility mechanism, but not ear nonlinearities. But it doesn't actually have anything to do with the time resolution of the music format itself - merely the frequency resolution. If one were able to bypass the ear's nonlinearities and somehow drive a DAC directly into the brain, Kunkur's result implies that all you really need for that is still 44khz. The justification for higher res relies entirely on intrinsic distortion.
Not at all. The only "real life application" here is listening to music, and humans don't "listen" to waveforms - they "listen" to a filterbank driven by waveforms (processed by the brain yadda yadda yadda). In such a context, combined with knowledge about the limits of hearing, whether or not the resolution advantages of eg DSD are actually important is entirely questionable.
Of course added sample rate and bit depth is useful when doing certain numeric things (like counting kids). Many scientific applications do benefit from increased performance like that. But digital audio is not necessarily one of those situations.
Kunkur says some about the ultrasonic content of certain instruments, the extremely rapid attack of things like xylophones, etc. And it is true that if you make a time resolution measurement out of the attack time of an instrument - say, measuring the time between the absence of an attack and the conclusion of an attack - 16/44's performance would be clearly worse than 5us (in fact it really would be somewhere in the 23us range). But that's only happening because you are measuring the time elapsed between the non-existence of an event to the existence of an event. You really don't have any other choice but to mark the start point on the exact sample positions, so obviously shortening the time between those sample positions is going to make the measurement more accurate. But if you time the start point based on an event rather than a non-event - ie based on an estimation of a location on the signal itself, similar to how the end point is dictated by the position of the peak value - the disadvantages of 16/44 suddenly disappear. The position of the "non-event" (the last time at which no event is observed) is completely arbitrary; the measurement changes depending on the subsample delay of the signal as a whole; the ear does not care one whit about the location of such samples. I say such a test is invalid.
It's not. That's precisely my point. It (and the entire pulse discussion) is strictly a didactic exercise to demonstrate what is being misconstrued about digital audio and time resolution. The ear does not count heads or measure pulse widths, and that alone is a rather valid argument demonstrating that the claimed superiority of DSD's performance is both mistaken and irrelevant.
More to the point, and very ironically: The audiophiles in this discussion are the ones arguing from predominantly objective grounds - that this or that numeric measurement shows the inferiority of 16/44. The skeptics in this discussion (me, jj, krab) are the ones arguing that such objective measurements are invalid, and that the actual act of listening will not back up the objective claims. One could *almost* go so far as to say that jj, krab and I are the ones being the truest to Stereophile's core philosophy here!
So, I actually did this, in the course of posting my questions in the other thread.
I do see that he might have a point regarding nonlinear mixing in the ear causing intermodulation down into audible bands, and I am sort of wondering if his test method is novel. It's possible that it could be used as a theoretical justification for high res.
There are still many, many questions that are worth asking as to his analysis methodology, as I point out. This isn't being overly unfair to him specifically - I raised similar questions about Meyer/Moran (in a rare case of me actually agreeing with Clark Johnsen for once).
It is still completely possible, even given Kunkur's results, that the audibility of high res may never be demonstrated for commercially produced music. Thus, high res audio would at best only be justified in an audiophile context, when theoretical faults are to be minimized, rather than observable faults. That situation would imply that much confidence should be placed in high res.
He's still wrong in all sorts of ways about digital audio, as I make abundantly clear.
Ultimately, whether or not a paper is peer reviewed doesn't really matter when it's wrong. Peer review is a demonstration of trust in the methodology by experts in the field, and does not preclude reasonable and scholarly later debate, of which some is actually happening here.
I mean... do you really want to make this argument an appeal to expert opinion? Fine. Kunkur has a PhD in physics and does not state his undergraduate degree on his resume. He does not appear to have any professional or academic training in digital audio or signal processing. I have an undergraduate degree in EE and have significant education on digital audio and signal processing (like any accredited institution provides). This topic largely revolves around digital audio and signal processing. Not about physics. This evidence suggests that I am, factually, more educated on this topic than he is. Perhaps not smarter, but almost certainly more educated.
So why should I consider him, or his peers, any more authoritative than myself? Why should I not rationally question his paper simply because it passed peer review?
'All this effort' amounted to one post from me about the Greiner article...not counting this one..and one JAES search that took seconds.
If you fail to find the article -- or any(!) article by Greiner -- using the AES search engine (the 'AES Journal Article Database Search' to be exact), you're doing something wrong...because I assure you it's there and easily found. Just follow my hint above. (I'll repeat an important clue for finding the exact Greiner article: search text in titles AND ABSTRACTS for the word 'cable")
Seriously, if you're going to claim I'm 'bullshitting' and it doesn't exist, you're going to end up looking pretty silly. You seem to be one of the more reasonable Stereophile forum denizens so I'll give you one more shot at finding it before I post a direct link. Heck, at this point I'll give you points just for finding ANY article in JAES by Greiner. There's twelve of them!
Oh for Jebus' sweet sake, if I'd known you were Scott Wheeler, I'd have not bothered at all. Your obtuseness is as well known to me as your long history of arguing with Arny on RAO and RAHE.
Anyway, here's the article you weren't competent to find, Scott:
http://www.aes.org/e-lib/browse.cfm?elib=3987
Amplifier-Loudspeaker Interfacing
JAES Volume 28 Issue 5 pp. 310-315; May 1980
ABSTRACT: Loudspeaker cables are investigated to determine if they exhibit transmission-line characteristics. Lumped representations of cables are suggested with consideration of the effects cable parameters might have on the audio signal being transmitted. Interaction of the cable with the amplifier, the loudspeaker, and other elements which may be part of the amplifier-cable-loudspeaker circuit are treated.
*Anyone* can find this article from the AES search engine link that I already posted.
http://www.aes.org/journal/search.cfm
Putting 'Greiner' in the Author box, with nothing else, returns 12 articles, the above one included. Greiner in Author + 'cable' in the Title/Abstract Search box, also brings up the article, *as long as you check 'Abstract' search*. Because, as you'll notice, the word cable isn't in the title.
Heh..it's always funny how that happens. Whenever 'subjectivists' try to marshal 'science' like that misleading Pyramix marketing slide, they decide to ignore the science of the thing they claim to put before all: hearing.
We bad guys on he other hand, have said more than once: we can measure stuff that can't be heard. It's why square wave and the Pyramix demos are bogus; with some further caveats it's why Atkinson's MP3 article and demos are bogus. In the particular case of Pyramix and square waves, the ear is a lowpass filter; so is all real-world audio recording and playback equipment. So some thing that 'measures' better or worse than another due to high-frequency content, will not necessarily sound any different.
Your post is so ridiculous I don't know whether to reply or not.
Actually I saw immediate attacks such as Dr. Kunkur does not understand basic sampling. That does not sound like debate.
Dr. Kunkur has credentials and a rigorous process. I see none from you nor from your side.
And who says Dr. Kunkur knows less and you more? Dr. Kunkur's describes the rigors which I am posting below.
Maybe because he is a professor. You comment appears very immature.
Sorry if this seems harsh, but you really have no credentials compared to him and he has all the rigors of the process, including experts.
For those who missed it, here are the rigors Dr. Kunkur had to "endure". Note: I broke up paragraphs for easier reading.
"For those who have no idea what science and the (incredibly rigorous) scientific process is, let me explain what went into publishing the two above mentioned papers that have apparently generated controversy among lay readers (but no controversy whatsoever in all the professional circles, which include audiolists, otolaryngologists, acousticians, engineers, and physicists ).
An experiment has to be carefully thought out and then submitted as a proposal to an Institutional Review Board (IRB) and approved by them before it can even begin. Then optimum equipment, methods, and a multitude of cross checks has to be developed (if you read my papers in their entirety, you will appreciated what went in).
The results, analysis, and conclusions are then carefully considered and discussed with colleagues who are experts in their related interdiscplinary fields; for this I went in person to various universities and research institutes and met with people in departments of physics, engineering, psychology, neuroscience, music, communications sciences, physiology, and materials science.
After that the results and conclusions were presented at conferences of the Acoustical Society of America (ASA),
Association of Research in Otolaryngology (ARO),
and American Physical Society (APS).
Seminars were also made at numerous universities and research/industrial institutions (please see the list on my web site). After each presentation, the audience is free to tear apart the conclusions and ask all possible questions.
Eminent people such as presidents of the above mentioned societies and corporations (ASA, ARO, Bose corporation, etc.) have been present during my presentations. After passing through this grueling oral presentation process, written manuscripts were then submitted to journals.
There, anonymous referees are free to attack the submission in any way they want. More than a dozen referees and editors have been involved in this journal refereeing process.
Only after everyone is satisfied with the accuracy of the results and all statements made in the manuscript, are the papers published in the journals. The entire process took around 5 years from initial concept to refereed publications."
Let me interject that all the PHDs, experts in their field must agree with the testing, conclusions etc, then referees before it can be published.
"By comparison exactly what are the credentials of the forumee who made this off-the-cuff claim about the temporal resolution depending reciprocally on the bits? What is the journal citation for this? Readers should be wary of obtaining their wisdom from web sites. When possible they should always refer to original journal articles (these should be refereed scientific journals not magazines).
I hope this clarifies the meaning of temporal resolution in the context of sound reproduction systems.
Sincerely,
Milind Kunchur
Scott,
In fairness to both sides, I took up the challenge to find the article as suggested by 'vocal representatives' from both factions of this issue, using the internet search. I found the organization AES and their website. I entered the author's name "Greiner" and the search immediately found the article regarding '...loudspeaker cables..." as "...transmission lines...", - end of story, IMO.
All this bickering from all those participants is really unnecessary, distracting, and detracting from this forum's purpose. We're discussing issues that are for the most part shades of gray, not black and white. Sampling theory, like many mathematical theory is very clear about what it can and cannot do, and not much gray there.
On the other hand, IRBs at an university really varies with who is on the IRB or not (more grayness). They are professionals with excellent credentials, and sometimes may not have the exact expertise in a given area. Both sides of the argument may have valid points here as had been stated by them, and those points are not mutually exclusive or far-fetched (eg. IRBs are knowledgable about the sampling theory field vs. they're concerned about the welfare of the experimental subjects -- as both could be true). Since I'm not at a university, but deal with them regarding IRBs, I know a bit about this.
Regarding the two viewpoints presented (from the Stereophile 'regulars', and from the Hydrogen forum 'regulars', I see no disjoint between the ideas presented. The former is talking about resolving two pulses in time through the usual sampling method used by the 16/44 system, whereas the latter is talking the ability to discern two pulses in two independent channels, one in each channel. I found the explanations understandable.
The good doctor from South Carolina is describing the situation advanced by the former group, not the what the latter group is talking about. If I recall correctly, "jj' pointed out this difference many pages ago.
Sorry, but I can't see how so many pages of back-and-forth is worthwhile, especially when understanding the technical points is missed. Can we get back to reading about the technical explanations, without the flaming? That was achieved some pages back in the middle of this sequence, and we apparently have lost it again.
What's clear is that Michelle J. Frogger and the others falsely posing as experts and saviors of the religion of the High End (which must be separated from those of us who like high-quality sound, please, OR those who prefer various euphonic effects, all of which pass under "preference") have no realization that the bandwidth of the widest filter in the human ear is about 6kHz wide, and that even inside of a bandwidth under 100 Hz (at 300Hz center frequency, for instance), one can resolve 50 microseonds in interaural time difference.
The fact is that the ear has been demonstrated, over and over again, to consist of a mechanically implimented filter bank, with bandpass filters, is in absolutely no serious discussion inside of science. The bandwidth of these filters is memorialized in a well-known standard, the "Bark" scale, and has been further refined (to somewhat smaller bandwidths) to the an "ERB", the scale of which is presently under debate.
These are both measuments of the ear's mechanical filterbank, and neither has a bandwidth even remotely approaching 20kHz, let alone 50kHz. The center frequencies are allocated in the range of 30Hz to approximately 16kHz, with bandwidths wider at high frequencies, approaching about 1/3 to 1/4 octave.
This has been true for years, actually.
And who do these organizations support?
Dr. Kunkur and his conclusions. Sorry.
jj,
What is the finest frequency resolution in the mid-band for the human ear, say around the 'concert A' of 440 Hz?
"What's clear is that Michelle J. Frogger and the others falsely posing as experts and saviors of the religion of the High End (which must be separated from those of us who like high-quality sound, please, OR those who prefer various euphonic effects, all of which pass under "preference") have no realization that the bandwidth of the widest filter in the human ear is about 6kHz wide, and that even inside of a bandwidth under 100 Hz (at 300Hz center frequency, for instance), one can resolve 50 microseonds in interaural time difference."
You might as well say the sky is blue. That makes about as much sense. You do win the Trifecta Award of the Day for the longest run-on conditional sentence containing the most logical fallacies, however.
Never said the article didn't exist. I simply asked for more info so I could find it. You and Arny made all these ridiculous accusations about me not really wanting to read it. *That* is "bullshit." And frankly if Arny really wanted me to read it all he had to do was email it to me. So, Steve, the "bullshit' is all the ad hominem based on Arny's and your "bullshit" presumptions about me. none of which has any substance when it comes to audio.
Well, if you had checked, the pyramix reproduces the 3us pulse as expected, very poorly. So no misleading graph there. Remember the ear is capable of detecting to at least 5us, as Dr. Kunkur has demonstrated. And remember, time constants accumulate in a system (total frequency response falls), and 16/44 is only one leg of the chain.
So by your, and other's, limited standards, 16/44 is good enough. Yet 16/44 represents the absolute minimum in terms of high frequency response, approximately 20khz, in the chain/multiple stages of a system. Any other properly designed component in the chain will have superior frequency response to 16/44. The truth is that as we add components to the system, the actual high frequency response dramatically falls off, so much less than 20khz. So by your own limited standards 16/44 is grossly insufficient.
I thought your goal, and your friends, was to help people obtain the best possible reproduction of music possible.
You are confusing two questions.
The filter in the ear will have a width of circa 80 Hz or so, depending on authority you choose (there are some who argue for narrower, but there is a major dispute about their numbers).
HOWEVER, the ability to detect pitch is far, far better than that, of course, because of the ability of the ear to detect the arrival of excitation above and below the center frequency of the filter, where the phase of the filter changes very rapidly in a very short amount of time.
This change in phase, also coupled with harmonics usually present in signals that reinforce the time-domain signal (bear in mind that under 500Hz the ear is very strongly a phase detector for the original signal, and that this behavior goes up to at least 2kHz at dimished levels), provide a frequency resolution of a few cents. I'd have to look up the actual number, but it's MUCH smaller than the bandwidth of any the cochlear filter.
Indeed, it clarifies things quite a bit. It would appear that Dr. Kunkur, in his insistance that the number of levels in a PCM signal does not affect the ability to resolve time of a signal arriving into the PCM system, has made a fundamental mistake. This, as always, from your quotes of his work.
There is no doubt about this, were he right, the modem in your cell phone, your HDTV, etc, would not work.
And they do work. Claiming that time resolution (and thereby phase resolution, which is nothing more than another measure of time) does not affect the ability to resolve time would require, among other things, that no modem could do better than 1 bit/Hz.
As many do, there really is no doubt. As quoted, this is a simple, and very frankly fundamental, mistake.
And, remember, that time resolution can be shown to be resolvable within a 44/16 system. So what's your point?
Actually, no. I would suspect you've heard of the term "dominant pole"? If all the other parts of the system are well above 20kHz at their 1dB cutoffs, then there is no doubt, the system response is going to be very, very close to the response of the 44/16 system.
Your claim otherwise is, well, ...
Easily tested, as well. Perhaps you should do so.
Which is why we examine actual, relevant, and well-known sources of impairment that have nothing to do with 44/16 systems. Sources of impairment that date back to Steinburg, Fletcher, and Snow, from the 1920's. KNOWN sources of impairment that are documented and published by the very people who originated the whole idea of electronically reproduced audio. Work from those people that stands to this day, and is known to be accurate and important, but that your present-day stereo system simply ignores.
That's what we work on, actual issues that make a huge difference, not small issues that may or may not matter in very constrained conditions with mistaken assertions about the sampling theorem and the time resolution of PCM systems.
[quoteThat makes about as much sense. You do win the Trifecta Award of the Day for the longest run-on conditional sentence containing the most logical fallacies, however.
I am hardly responsible for your inability to understand good english. Got any morphic fields to play baseball on?
[quote}
Never said the article didn't exist.
You're lying of delusional, Scott. This whole interchange started when you said quite exactly that the article didn't exist.
Furthermore, I just entered "Greiner" (clipped from your post) and plugged it into the JAES search engine.
I got numerous hits, which contradicts your claims.
The first hit says "Loudspeaker cables are investigated..."
Scott, that's three strikes against you.
Had you signed your posts with your actual name Scott, I would have known that you are the poor delusional soul who has been entertaining us with your crazy antics on RAHE.
I would have also known that you are the poor delusional soul who tried to sue me for libel in California Superior Court over a post on Usenet, but had your case thrown out of court.
Pages