In audio I personally care of power gain; however with non-linearities in mind, from other side. Especially dealing with SS devices you can't get both at once.
I never tried to load infinitely high output on infinitely low input; however I used similar things deliberately creating guitar effects, but never tried for a high-end audio. How good are OL parameters of such a thingy? You should know better than me...
I will, once again say: it might be the result of PIM, but might be not. I did not follow this thread carefully when assumed that you have a proof because of asymmetrical sidebands, but you denied that, saying that sidebands are equal. That means, no proof at all.
Non-harmonic content can't prove presence of PIM. It may prove presence some say thermally-induced oscillations that cause something not related to signal spectra, or harmonic content may look like non-harmonic when intermodulaions occur between periodic signals and their envelopes (discussed with Jan Didden), but no way non-harmonic content can signify presence of phase modulation.
"Amicus est Socrates, magister meus, sed magis est arnica veritas." (C)
Last edited:
Hi John,
I really do think you need to read other people's responses more carefully.
Many people have now pointed out the normal and proper way to investigate any phenomenon. You discover something that you make note of. You recreate the observation, and once you can do that reliably, you make sure you don't have an error in the experiment. Once you are sure no errors have been made, you invite peer review and perhaps write a paper on your work. This would include the exact test setup, stimuli (test signals) and any other pertinent information so that your observations can be replicated by others. In this case, you want to see other experimenters come up with the same data. At this point, conclusions are not important. It's the validity of your experiment that is the critical question.
"If it looks like a duck, quacks like a duck, maybe it is a duck, or just possibly it's a decoy, a smudge on your glasses or whatever else was in the bottle along with the alcohol".
Note that I'm not suggesting that you drink, just looking at a better comparison. Now, you say it was a duck. Did you manage to get a picture so we can all have a look?
As said before by many other members, we can accept things exist that can not be measured at this time. No argument there, so you can drop that issue.
More than one member has expressed interest in running your experiments as close to the original set as possible. The hope is that others can recreate what you were seeing. This is not an attack of any kind. This is called "peer review", and this is part of any technical announcement. You can not expect for a moment that anyone will simply drop what they know in order to accept what you are telling them without any experimental data or other proof. You wouldn't, but you are expecting everyone else to do exactly that. Can you see why you are being questioned about these things John?
Now, for the benefit of all members, would you please draw out your experiment exactly, along with signal levels and any other information that might be required? This would go a long way to unifying everyone here and get them onto your page. Then you can talk about these things in a meaningful way.
-Chris
I really do think you need to read other people's responses more carefully.
Many people have now pointed out the normal and proper way to investigate any phenomenon. You discover something that you make note of. You recreate the observation, and once you can do that reliably, you make sure you don't have an error in the experiment. Once you are sure no errors have been made, you invite peer review and perhaps write a paper on your work. This would include the exact test setup, stimuli (test signals) and any other pertinent information so that your observations can be replicated by others. In this case, you want to see other experimenters come up with the same data. At this point, conclusions are not important. It's the validity of your experiment that is the critical question.
Thank you. I had asked if that diagram accurately represented your test setup a long time ago and got no response from you.
That is a twisted analogy you used. It should read more like this ....
"If it looks like a duck, quacks like a duck, maybe it is a duck, or just possibly it's a decoy, a smudge on your glasses or whatever else was in the bottle along with the alcohol".
Note that I'm not suggesting that you drink, just looking at a better comparison. Now, you say it was a duck. Did you manage to get a picture so we can all have a look?
As said before by many other members, we can accept things exist that can not be measured at this time. No argument there, so you can drop that issue.
More than one member has expressed interest in running your experiments as close to the original set as possible. The hope is that others can recreate what you were seeing. This is not an attack of any kind. This is called "peer review", and this is part of any technical announcement. You can not expect for a moment that anyone will simply drop what they know in order to accept what you are telling them without any experimental data or other proof. You wouldn't, but you are expecting everyone else to do exactly that. Can you see why you are being questioned about these things John?
Now, for the benefit of all members, would you please draw out your experiment exactly, along with signal levels and any other information that might be required? This would go a long way to unifying everyone here and get them onto your page. Then you can talk about these things in a meaningful way.
-Chris
The point I was making about the reference to the piano tuning was that a harmonically correct distortion that should sound like normal instrument harmonics may actually not correspond to the real sound and to a sensitive listener may sound "wrong". The point of this observation is that we can't accept any generalization on the audibility of distortions or lack thereof. The assumptions need to be tested.
As for linear distortions I belive it is accepted that very small frequency response errors, on the order of .1 dB over a large range, can be heard. However small peaks and dips can be very hard to discern. I know from experience small changes in RIAA eq can have a major effect on the perceived sound, but sonically don't match what the actual change was.
As for linear distortions I belive it is accepted that very small frequency response errors, on the order of .1 dB over a large range, can be heard. However small peaks and dips can be very hard to discern. I know from experience small changes in RIAA eq can have a major effect on the perceived sound, but sonically don't match what the actual change was.
Chris:
http://www.linearaudio.nl/Miscellaneous/langevin im in tape recorders.pdf
I had a cursory look and it appears that the paper addresses all those 'unexplained' spectral lines in the famous Fig 3.
jd
Right, thanks. Although the effect can be minimized by the same measures that minimize AIM and THD, some applications can be more sensitive to it, for instance because of sensitivity to slew rate limiting. An example would be an I/V converter where very high input slew rates are encountered. In this case you can have relatively large OL gain and phase modulation which, as Hawksford maintains, could even be considered as the analog equivalent of jitter - and sound similarly.
jd
Joshua,
You are missing the point of these discussions massively. It is not about whether science has answers or not. It is not about whether PIM or TIM exists or can be audible. That is not contested as such.
The discussion is about answering questions about a statement that is made, about providing arguments that support your position so that others can judge for themselves instead being limited to 'believe' or 'not to believe'. Providing full details of your measurements, so that others can repeat them and either confirm or falsify them.
In other words, intellectual honesty.
jd
On the other hand, if we measure something in a less traditional way, like RF contamination, it is marked as unimportant. "Show me the influence in audioband (under 20kHz)", is the answer, of course by spectrum analysis. We are permanently forced to prove by routine approach. If we quit routine approach, the results are unimportant or inaudible. Stage 2.
Re null test
The null test must be performed on a complete system. I.e., it has no value to test e.g. a link stage only. Different HF content on outputs of 2 link stages may have great influence on power amp behaviour. So do cables. I encourage anyone to perform a null test on a complete audio chain, with power amps and speakers. I can imagine excuses that it is difficult to perform.
The null test must be performed on a complete system. I.e., it has no value to test e.g. a link stage only. Different HF content on outputs of 2 link stages may have great influence on power amp behaviour. So do cables. I encourage anyone to perform a null test on a complete audio chain, with power amps and speakers. I can imagine excuses that it is difficult to perform.
piano tuning, inharmonicity
Hi Demian,
Under the assumption that you are talking about 'stretching', I'm afraid you missed the point. It's just the opposite. Inharmonicity -instead of harmonicity- is percieved as wrong.
'The tuner will stretch the notes, slightly sharpening the high notes and lowering the low notes so that the overtones of low notes have the same frequency as the fundamentals of high notes.'
In other words, the purpose of stretching is to compensate for (inevitable) inharmonicity of the strings.
Source: http://en.wikipedia.org/wiki/Piano_acoustics
Cheers,
Edmond.
Hi Demian,
Under the assumption that you are talking about 'stretching', I'm afraid you missed the point. It's just the opposite. Inharmonicity -instead of harmonicity- is percieved as wrong.
'The tuner will stretch the notes, slightly sharpening the high notes and lowering the low notes so that the overtones of low notes have the same frequency as the fundamentals of high notes.'
In other words, the purpose of stretching is to compensate for (inevitable) inharmonicity of the strings.
Source: http://en.wikipedia.org/wiki/Piano_acoustics
Cheers,
Edmond.
Did you do it?
The normal/logical approach in experimental sciences is:
1. You provide the evidence and all required information to reproduce your results. Be prepared for a critical approach and to answer questions.
2. Others are doing their best to reproduce your results. It may be required to answer more questions.
3. You get the confirmation or infirmation of your original results. Theoreticians will jump in, trying to find an explanation for the results.
Expecting somebody to invest and check what you think is relevant will usually not pass the ** filters, unless you have a very solid theoretical background for your assumptions and allegations (ala "the Higgs boson" of hifi audio).
Last edited:
Thank you.
Sometimes, certain types of distortion are being heard by some people, though not by all people, and it isn't completely set how to measure it. This have nothing to do with intellectual honesty.
- Status
- Not open for further replies.