@ Andy_C,
Of course, but on the other hand, as you know, a peer review can not assure that the content of an article is right, but should assure that the methodoly used meets scientific standards.
Obvious errors in mathematics for example will be detected and logical errors in the reasoning as well (hopefully 🙂 ), but no reviewer can duplicate any experiments.
And at least within the peer reviews on the JAES there are different sets of rules in respect to the focus of an article. An engineering report isn´t afair treated in the same manner like an scientific study.
So in the end i think theres not much difference, whether some dbt results were presented within an JAES article or just described in a conference paper. Every author deserves and gets the presumption that he describes his results in a honest manner.
And especially any experiment has to be successfully reproduced by others to get accepted as part of the standard of knowledge. That is the demand on every test- it has to be objective, reliable and valid.
Could you elaborate about it?
I had a very opposite impression but that might be influenced by other readings like in the pgm group for example.
Wishes
Of course, but on the other hand, as you know, a peer review can not assure that the content of an article is right, but should assure that the methodoly used meets scientific standards.
Obvious errors in mathematics for example will be detected and logical errors in the reasoning as well (hopefully 🙂 ), but no reviewer can duplicate any experiments.
And at least within the peer reviews on the JAES there are different sets of rules in respect to the focus of an article. An engineering report isn´t afair treated in the same manner like an scientific study.
So in the end i think theres not much difference, whether some dbt results were presented within an JAES article or just described in a conference paper. Every author deserves and gets the presumption that he describes his results in a honest manner.
And especially any experiment has to be successfully reproduced by others to get accepted as part of the standard of knowledge. That is the demand on every test- it has to be objective, reliable and valid.
Could you elaborate about it?
I had a very opposite impression but that might be influenced by other readings like in the pgm group for example.
Wishes
Just been adjusting my speakers to new and higher levels
Afterwards it came to my mind that the flawed adjustments I also made did stay in my ears fore a while after, making it a bit difficult to decide which was the good one...but I know my system enough to make the choise
Then it suddenly struck me...what about blindtests, are they "polluted" in the same manner, with testpersons not being able to choose one over the other because they are presented to both amps on short timebasis
BUT, then I also remembered that my friend and I very quickly chose my amp to be the best, with a wide margin...although he would have preferred his own newly bought amp to be the better one...now every time he looks at my DIY amps it with awe(?)
You see, despite of placebo and psycoacoustics...it was not that hard at all
Have you ever even heard a good amp...doesnt sound like it
YES, trust your senses...its the most important you have in this life
To those who dont trust or even use their senses, I only feel sorry fore them
If you were some innocent kids(could sound like you are) I wouldnt be bothered and leave it well alone, but as you are so agressive about it, it has to be commented
Afterwards it came to my mind that the flawed adjustments I also made did stay in my ears fore a while after, making it a bit difficult to decide which was the good one...but I know my system enough to make the choise
Then it suddenly struck me...what about blindtests, are they "polluted" in the same manner, with testpersons not being able to choose one over the other because they are presented to both amps on short timebasis
BUT, then I also remembered that my friend and I very quickly chose my amp to be the best, with a wide margin...although he would have preferred his own newly bought amp to be the better one...now every time he looks at my DIY amps it with awe(?)
You see, despite of placebo and psycoacoustics...it was not that hard at all
Have you ever even heard a good amp...doesnt sound like it
YES, trust your senses...its the most important you have in this life
To those who dont trust or even use their senses, I only feel sorry fore them
If you were some innocent kids(could sound like you are) I wouldnt be bothered and leave it well alone, but as you are so agressive about it, it has to be commented
tinitus
sure if you A/B two amps (or whatever components) it is possible to choose one that is inferior because you are used to it or becasue it fits the other (inferior) components in the rig. That's why it's clever to test transmission links in before/after listening tests (is this also called bypass tests?).
You insert the DUT in a complete playback (or recording) chain. If the signal is colored by the DUT you will hear it and also hear in what way it colors. If it can not be detected it's a good performer in that set up at least. You can never say a component is 100% transparent (which is something not even everybody is aiming for, I do though) but you can conclude that in this set up, in this room and with these recordings and with.. yada yada.. the DUT/component is free from audible errors.
This method is not perfect, but pretty close and actually the best way to test amps and preamps I'd say.
IMO it's a laugh that not more people (no matter if DIY or pro) use this method as a way of evaluating electronics. I'm about to set up my own test equipment so I can stop laughing at myself at lest. 😀
/Peter
sure if you A/B two amps (or whatever components) it is possible to choose one that is inferior because you are used to it or becasue it fits the other (inferior) components in the rig. That's why it's clever to test transmission links in before/after listening tests (is this also called bypass tests?).
You insert the DUT in a complete playback (or recording) chain. If the signal is colored by the DUT you will hear it and also hear in what way it colors. If it can not be detected it's a good performer in that set up at least. You can never say a component is 100% transparent (which is something not even everybody is aiming for, I do though) but you can conclude that in this set up, in this room and with these recordings and with.. yada yada.. the DUT/component is free from audible errors.
This method is not perfect, but pretty close and actually the best way to test amps and preamps I'd say.
IMO it's a laugh that not more people (no matter if DIY or pro) use this method as a way of evaluating electronics. I'm about to set up my own test equipment so I can stop laughing at myself at lest. 😀
/Peter
Jakob2 said:
Exactly! Therein lies my objection to the article.
As I see it, papers that present the results of mathematical analysis are different in some ways from those presenting results of statistical studies of listening tests. With the exception of references, the mathematical analysis in the former is self-contained, and any reader familiar with the engineering and physics of the situation, and the references, can walk through the equations and follow through to the conclusions. The mathematics can speak for itself without supporting data.
With statistical studies, if there are no supporting data, descriptions of test setup and so forth, the only things really left are the conclusions. In other words, the very things that are needed to reproduce the experiment as you mentioned above are missing in such a case. That is my objection to the article. It was structured around a set of anecdotes and unsupported assertions.
If I were asked to peer review such an article, I'd ask things such as "Can you supply the data?", and "What was the test setup?". I don't think such questions are at all unreasonable, yet no information of this sort was provided in the article.
Imagine if Neville Thiele had said "My analysis shows that the frequency response of a vented box loudspeaker has the form of a fourth-order high-pass filter" without providing any supporting analysis. It's not unreasonable to ask, "So, show me how that works!". Of course, in reality Thiele did a superb job of showing just that.
andy_c[/i] [B] After reading the article said:
Okay. Because the article consisted largely of anecdotes and unsupported assertions, I got the impression that the author was just blowing smoke. I realize that's not being objective, and that's why I left out this comment in my previous post.
What about Dr Geddes ? He is perhaps (with his wife) world's most authoritative subject on audibility of different types of distortions. Certainly one of the few people who understands correlation between metrics and subjective perception better than anyone in this thread.
Check what kind of amplification he frequently uses to demo his Summa.
Bratislav said:
We may not agree with him on his methodology or his choise (i won't comment on the latter as i haven't heard it -- there are some real surprises out there) but he chose the amp he did because it performed particularily well on a new distortion test he is developing, and he figured it performed better than other candidates.
dave
Ahem ... can I offer this direct quote from Earl :
"Literally any amplifier works fine.
Earl Geddes"
"Literally any amplifier works fine.
Earl Geddes"
planet10 said:
Please refresh my memeory, Dave. Was that Halcro ? Ongaku ? Wavac ?
Or was it something much more mundane, like lowly receiver from Technics/Onkyo ?
Bratislav said:
IIRC it was a Pioneer... he may say any amp will do, but that he went to the effort to develop a test for amplifiers, and then used it to choose an amplifier belies that statement AFAIC.
dave
john curl said:
I for one am grateful that Walker, Dunlavy, Linkwitz et al "wasted their time" in producing some of the world's best loudspeakers (truly pushing the boundaries of what is possible).
It is so funny that they all believed that amplification is essentially "good enough".
"Not trusting their ears" ? "Not critical enough" ?
I don't think so.
Loudspeaker designers often belittle the efforts of amp designers. Their problems are so great and easily measurable, that when something 'measures' pretty well, they consider it near perfect. I know this from experience.
This is hardly a worthy reply to janneman's thoughtful post.john curl said:
No. Only a digital link can be bit perfect. The analog audio after D/A conversion cannot be bit perfect, because there are no bits in analog. Any DAC or sound card that accepts the sampling rate you're sending it will get bit perfect digital data if you bypass the processing Windows normally does to audio--that's a software issue and covered by the player software I mentioned. The other issue is whether a setup will reject jitter at the interface between the PC and the DAC board, and thus jitter would only be dependent on the local DAC clock. Essentially all PCI sound cards do this and use an asynchronous connection--their DACs are timed by their local clocks and the PC fills their buffers as they get low. But an internal sound card is in the PC case, a very electrically noisy environment. External DACs, not only the Wavelength have an asynchronous connection. The EMU 0404 USB does as well, but I don't know if that can be considered an audiophile DAC--I guess you could mod it. Another option is to use the new ESS DAC chip which uses a digital PLL to recover an ultra low jitter signal.AndrewT said:
That sounds like a conspiracy theory... so is Hawksford in on it too? I mean, he has lots of papers in the JAES.john curl said:
No, you also have a mind with the ability to reason. Or, at least some people do...Andre Visser said:
Well, it's much harder to design low distortion speakers than low distortion amplifiers, so this situation doesn't seem surprising. But this is a criticism of both sides: perhaps if speaker designers made their speakers not so terribly high in distortion, differences between electronics would be easier to hear.Bratislav said:
we're getting closer to plain english.abzug said:
Now that I am starting to understand some of the content, I'm beginning to doubt some of your contentious claims for bit perfection and why it could/could not be achievable.
And what claims are these? Bit perfection is really only an issue when using PCs as source, and is a software issue. I haven't made any contentious claims about it, nor have I said it's not achievable. It's achievable any time you bypass the signal processing that the operating system does. I think you're confusing in your mind my comments on bit perfection with the issue of interface jitter rejection, which is completely separate.
john curl said:
Aren't "their" problem ours too ? I mean what do you use to listen to amplifiers ? Speakers I presume ?
Besides Peter Walker was amplifier designer too, not just speaker designer. He understood both sides of coin pretty well.
John Dunlavy was electronics engineer specialising in RF design before embarking on speaker design. He knew a thing or two about amplification.
And I guess Sigfried's doesn't need introduction. There's little doubt that he knew about electronics quite a bit, being a senor R&D engineer at HP for decades before retiring. Designing 20GHz+ bandwidth spectrum analysers in late 70's could have been a bit more challenging than a mundane audio amplifier ...
Hi Abzug,
I don't have the knowledge to argue the case. But the contentious claim is
That was why I gave you the get out of jail free card (apple).
I don't have the knowledge to argue the case. But the contentious claim is
go read the DIYhiFi thread on bit perfection and in there the software writers and hardware designers are discussing whether PCs can ever be bit perfect.abzug said:
That was why I gave you the get out of jail free card (apple).
- Status
- Not open for further replies.