Home |
Search |
Today's Posts |
#361
|
|||
|
|||
Bob Cain wrote: Also, a single tone cannot produce Doppler distortion so I am definitely wrong about that. Yikes! I see why. The definition I've been bandying about for a linear system is actually the definiton of a linear, time invariant system. The system we are considering must be time variant in terms of the impedences involved. This is getting wierd. Yikes is right. I'm not sure that thought would have survived a nights sleep but I hit Send instead of Save at bedtime. Anyway, that's what a raw speculation that needs much further thought looks like when it pops outa my head unbeckoned. If it's nonsense I appologize for the noise. I'm trying to understand why a system that appears linear in the case of any pure sinusoid would produce mixing when presented with superpositions. That defies my understanding at the moment but I plan on fixing that. :-) Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#362
|
|||
|
|||
In alt.music.home-studio,rec.audio.tech,rec.audio.pro, "Phil Allison"
wrote: ** The lack of the facility to post sketches and diagrams on usenet is a *real* drawback. When I need to explain stuff to non-technical folk ( and some technical ones too) I often reach for my pen and paper !!!! Draw the pics on paper, scan them in, put the files (reasonably compressed .jpg, please) on a website, and post the URL's in your message. Then, on second thoughts, the sketches that might appear most often could be kinda pornographic in nature ;-) Use Geocities, they're slow to clean up smut on their server. .......... Phil ----- http://mindspring.com/~benbradley |
#363
|
|||
|
|||
Mark wrote: Bob, You seem to think the Doppler effect doesn't happen in speakers beacue the air is moving with the speaker cone. I think this is wrong. The Doppler effect happens anyway. The Doppler effect depends only on the distnace between the Rx and Tx changing. But it does depend on the way that is changing. It does not occur for the reasons believed and it does not occur for at least one configuration that the vernacular theory says it will. Doppler happens for radio and light waves as well and there is no ether to move or not move. The Doppler effect is a function of the changing distance between the Rx and Tx and has nothing to do with the propogating medium. It has all to do with it. One way of stating why it doesn't occur in an infinite tube (or one terminated by an acoustic resistance equal to the characteritic impedence of air) is that in that case the radiation impedence seen by the piston exactly equals the characteristic impedence of the air. This simple fact guarantees perfect reproduction and no mixing of frequencies that will produce new ones. The vernacular theory about little HF waves being frequency shifted by big LF ones totally fails in this case. That's not at all what is happening. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#364
|
|||
|
|||
Bob Cain wrote in message ...
Gary is one nasty, sociopathic SOB.... You mindless and stupid assertions regarding the existence of dynamic Doppler distortion has been proven incorrect. Obviously the above remark reflects a case of sour grapes. It is also a classic example of the calling the kettle black. |
#365
|
|||
|
|||
Bob Cain wrote in message ...
Gary is one nasty, sociopathic SOB.... You mindless and stupid assertions regarding the existence of dynamic Doppler distortion has been proven incorrect. Obviously the above remark reflects a case of sour grapes. It is also a classic example of the calling the kettle black. |
#366
|
|||
|
|||
Bob Cain wrote in message ...
Gary is one nasty, sociopathic SOB.... You mindless and stupid assertions regarding the existence of dynamic Doppler distortion has been proven incorrect. Obviously the above remark reflects a case of sour grapes. It is also a classic example of the calling the kettle black. |
#367
|
|||
|
|||
Bob Cain wrote in message ...
Gary is one nasty, sociopathic SOB.... You mindless and stupid assertions regarding the existence of dynamic Doppler distortion has been proven incorrect. Obviously the above remark reflects a case of sour grapes. It is also a classic example of the calling the kettle black. |
#368
|
|||
|
|||
Scott Dorsey wrote: So I think folks should continue investigating doppler distortion because it's an interesting problem even if not a terribly important one. Mostly agreed. The extent to which unbelievably small effects are claimed to be audible on the ProAudio mailing list, if they are other than imagination, does push any such effects like we are talking about into an arena at least worth discussing, if not important. It wouldn't take much _at all_ to swamp the things they consider very important. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#369
|
|||
|
|||
Bob,
What makes you think Doppler does not occur in an infinite terminated tube. This is a similar case to a sliding RF transmission line. A radio wave in a transmission line that is changing length will experience Doppler shift. A radio wave transmitted by a perfectly matched antenna will experience Doppler shift if the antenna is moving relative to the receiving antenna. In fact, the receiving antenna can be the one that is moving. If the train is standing still, and you drive past in a car, the train whistle will still Doppler in your ear anyway. Doppler has nothing to do with impedance matching or anything like that. Shift happens :-) I think you are stuck on the idea of linearity. FM is inherently a non-linear process in the mathematical sense. If you take a perfectly linear VCO (in the sense that the frequency is exactly proportional to the control voltage) , it will still create sidebands around the carrier because FM itself is a non-linear process. So in the speaker case, even if the cone and the air are perfectly linear loaded and the 2 waves add perfectly in amplitude, the Doppler that produces FM is still a non-linear process and produced new sidebands. Non-linear does not have to imply a defect of some kind. Perfect FM modulators are non-linear in the sense that the modulation process creates new sidebands. Now in a real speaker which also has amplitude non-linearities due to changing spider tension etc etc. it is defiantly a debatable issue as to weather or not the Doppler effect is SIGNIFICANT relative to all the other non-linear effects going on. I have no idea. Mark |
#370
|
|||
|
|||
Bob Cain writes:
Randy Yates wrote: Bob Cain writes: Randy Yates wrote: So it's dropped. Randy, if you are going to be so nasty, the least you could do is refute what I've said. Let me get something out of it anyway. How can I refute what you have refused to clearly restate? All right, I'll accept that you haven't read the thread. I've found a new way to state it anyway, that I think makes it clearer. I'm saying that the instantaneous velocity of the piston is transfered to the wave in the right position, since it is moving in step with it, to propegate that velocity out as an acoustic wave. It is _in_ the acoustic wave it is creating and it is at the right place at all times to impart the correct velocity _because_ it is in it. It doesn't matter what signals might have been mixed upstream to get the signal that controls the velocity of that piston. It will be moving in lock step with the wave defined by that signal and will always be in the right place to deliver the right velocity to the outgoing wave. If, on the other hand the piston is moving with a constant velocity superimposed on the signal velocity, it has no way to transfer that constant velocity to the air because contant velocity doesn't create a wave. At f=0 it runs out of punch. It ceases abruptly to transduce at all. In that case, the piston will always be in the wrong position to correctly impart the desired velocity signal and that error is Doppler shift. If the above is truly crazy, then I'll agree that I'm over the edge. But I don't think so. What, exactly, is wrong with it? I don't want to talk about trains and whistles. Refutations: a) Non-Causality: If the speaker cone and the wave are EXACTLY in sync, then the wave must instanteously "know" which way the piston is going. This is ludicrous. b) Discontinuous Frequency Offsets: Doppler operating as you describe would result in a sharp discontinuity in the perceived frequency as the relative acceleration changed from zero to non-zero. Similarly for the opposite case - i.e., while the acceleration is non-zero there would be no Doppler shift but as soon as it reaches zero the Doppler would discontinuously jump to the frequency predicted by the Doppler equation. This is ludicrous. c) Acoustic wave propagation in air is longitudinal and thus pressure-waves. The velocity of the transducer is not related to the velocity of the wave and imparts no information to the propagated signal. -- % Randy Yates % "I met someone who looks alot like you, %% Fuquay-Varina, NC % she does the things you do, %%% 919-577-9882 % but she is an IBM." %%%% % 'Yours Truly, 2095', *Time*, ELO http://home.earthlink.net/~yatescr |
#371
|
|||
|
|||
Methinks The Ghost has a compulsive need to repeat everything three times.
Maybe he labors under the misapprehension that repeating a falsehood enough will make it true. "The Ghost" wrote in message om... Bob Cain wrote in message ... Gary is one nasty, sociopathic SOB.... You mindless and stupid assertions regarding the existence of dynamic Doppler distortion has been proven incorrect. Obviously the above remark reflects a case of sour grapes. It is also a classic example of the calling the kettle black. |
#372
|
|||
|
|||
If you want believable, try it for yourself and compare the results
across a wide range of LF and HF tones. I suspect it's because we're dealing with digital waves, but there could be other factors involved. "Arny Krueger" wrote in message news "Porky" wrote in message "Arny Krueger" wrote in message ... "Porky" wrote in message The experiment I suggested will give the results I gave, and if it is right at under the circumstances I suggested, it should be right under all circumstances with the same conditions, right? In other words, if it applies with a LF of .1 Hz or 1 Hz, it will still apply at LF's 20Hz or 50Hz, is that not correct? right. However, its a lot harder to properly measure doppler when the LF tone has a very low frequency. To measure it with a FFT you must use a FFT size that covers at least one cycle, and hopefully several cycles of the process. If the LF tone is 0.1 Hz, this means an absolute minimum of 10 seconds of data, and ideally 30 or more. At 44,100 Hz sampling, this would be a FFT composed of a minimum of 441,000 samples, and preferably several million samples. Consider the original example - the LF tone was 50 Hz. It had an 882 sample period. Note how much overkill there was when analyzed using a 65k sample FFT, or as I used a one million point FFT. One of the problems with FFT analysis that we've all overlooked is that we aren't really dealing with analog waveforms in our simulations, and we can get erroneous results when using high FFT numbers because we start playing in the digital "cracks", so to speak, I'd like to see a believable fuirther explantion of that. FFT and I are going on our 42nd year, and we've been pretty good friends the whole time. Having done a bit of experimentation, I've found that I get the most consistent results across the whole range be using an FFT number of 16K or 32K, higher rates give false results, especially at higher frequency HF tones. I'd like to see a believable further explanation of that. Alternatively, if your equipment will handle it, try creating the wave models at 24 or 32/96, or even 32/192, you'll see a considerable difference in your results, especially at higher HF tones and FFT numbers, and your results will be more consistent across the entire range of LF and HF tones. I'd also like to see a believable further explanation of that. |
#373
|
|||
|
|||
When dealing with analog, I would agree with you wholeheartedly, but since
any digital is only an approximation of the real thing, even with a very close approximation there is an error factor that increases as the tone frequency to sample frequency ratio decreases, (as the tone frequency gets higher. Doing tests that were identical in every respect, except that one test wave was 16/44.1 and the other test wave was 16/48, resulted in quite different analyzer waveforms. This is what suggested to me that it might be a digital issue. "William Sommerwerck" wrote in message ... One of the problems with FFT analysis that we've all overlooked is that we aren't really dealing with analog waveforms in our simulations, and we can get erroneous results when using high FFT numbers because we start playing in the digital "cracks", so to speak, I have to agree with Arny on this. (I used to do FFT and waterfall measurements when I reviewed for Stereophile.) Higher sampling rates are almost always better, other than their effect on measuring LF response. Regardless, the higher the rate, the _fewer_ the cracks. |
#374
|
|||
|
|||
Actually, using a "brick" analogy, digital may be bricks of varying widths,
but analog is one solid sheet of very smooth concrete, and even the very best laid brickwork still has cracks. "Arny Krueger" wrote in message ... "William Sommerwerck" wrote in message One of the problems with FFT analysis that we've all overlooked is that we aren't really dealing with analog waveforms in our simulations, and we can get erroneous results when using high FFT numbers because we start playing in the digital "cracks", so to speak, I have to agree with Arny on this. (I used to do FFT and waterfall measurements when I reviewed for Stereophile.) Higher sampling rates are almost always better, other than their effect on measuring LF response. Regardless, the higher the rate, the _fewer_ the cracks. Strictly speaking there are no cracks, its just that the bricks are wider. |
#375
|
|||
|
|||
"Randy Yates" wrote in message ... "Porky" writes: "Arny Krueger" wrote in message ... "Porky" wrote in message The experiment I suggested will give the results I gave, and if it is right at under the circumstances I suggested, it should be right under all circumstances with the same conditions, right? In other words, if it applies with a LF of .1 Hz or 1 Hz, it will still apply at LF's 20Hz or 50Hz, is that not correct? right. However, its a lot harder to properly measure doppler when the LF tone has a very low frequency. To measure it with a FFT you must use a FFT size that covers at least one cycle, and hopefully several cycles of the process. If the LF tone is 0.1 Hz, this means an absolute minimum of 10 seconds of data, and ideally 30 or more. At 44,100 Hz sampling, this would be a FFT composed of a minimum of 441,000 samples, and preferably several million samples. Consider the original example - the LF tone was 50 Hz. It had an 882 sample period. Note how much overkill there was when analyzed using a 65k sample FFT, or as I used a one million point FFT. One of the problems with FFT analysis that we've all overlooked is that we aren't really dealing with analog waveforms in our simulations, and we can get erroneous results when using high FFT numbers because we start playing in the digital "cracks", so to speak, The waveform being analog or digital makes no difference as long as sufficient bandwidth and dynamic range has been supplied by the A/D conversion. Rather, the problem you are ignorantly referring to is that an FFT implicitly assumes the input is periodic. If it isn't, you can get yourself befuddled. There is also the problem with using the FFT to estimate the spectrum of a random signal - it can be shown that there will be variance in the frequency estimates no matter how many points are used in the FFT (see, for example, "Signal Processing: Discrete Spectral Analysis, Detection, and Estimation," Mischa Schwartz and Leonard Shaw). -- Then why would the analysis show differing results with identical waveforms whose only difference was that one was 16/44.1 and the other was 16/48? Note that the difference was considerable. Since all the wave simulations were done with Cool Edit, could it be something in the way Cool Edit generates the sine waves? |
#376
|
|||
|
|||
Randy Yates wrote: Refutations: a) Non-Causality: If the speaker cone and the wave are EXACTLY in sync, then the wave must instanteously "know" which way the piston is going. This is ludicrous. No, total locality. The air is in contact with the piston. It knows it intimately and passes the information on down the line. In the tube, the radiation impedence seen by the piston is identically the characteristic impedence of the air. Consider the implications of that. I no longer believe that Doppler distortion is non-existant, simply that the vernacular justification, the one reiterated by Seigfried, is totally wrong. It is much more complicated than that simple principle. It will only exist in the far field where the transfer function from source to reciever varies with frequency and to the degree that it does. In the tube that transfer function is a constant, real function of frequency for as far as you want to go from the piston, i.e. an acoustic resistance. More on all that to come. It's taken me a while to figure out what's really going on and as of today I have the assistance of Art Ludwig who has not considered this phenomenon before. So far, at least, we seem in agreement. I hope with his help to arrive at the full expression of it, at least on the axis of a piston in a baffle, that I've been looking for. FWIW, the full expression of it is entirely dependant on the physical configuration of the speaker and how it is thus coupled to the air as well as the position in space from which the phenomenon is measured. 'Taint simple ay'tall. b) Discontinuous Frequency Offsets: Doppler operating as you describe would result in a sharp discontinuity in the perceived frequency as the relative acceleration changed from zero to non-zero. Why? Similarly for the opposite case - i.e., while the acceleration is non-zero there would be no Doppler shift but as soon as it reaches zero the Doppler would discontinuously jump to the frequency predicted by the Doppler equation. This is ludicrous. I'm not following this at all. Where is there any discontinuity here? The signals we have been discussing are infinitely differentiable. c) Acoustic wave propagation in air is longitudinal and thus pressure-waves. The velocity of the transducer is not related to the velocity of the wave and imparts no information to the propagated signal. It's not related to the rate of propegation but it is very related to the particle velocity that propegates within the wave. Randy, I really do appreciate the specificity of your argument. It's been really hard in this discussion to get ahold of anything with the generalities that have greeted everything I've tried to pin down. The understanding I'm arriving at would have come _much_ sooner with attack such as yours, correct or incorrect. Thanks, Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#377
|
|||
|
|||
"Porky" wrote in message
Actually, using a "brick" analogy, digital may be bricks of varying widths, but analog is one solid sheet of very smooth concrete, and even the very best laid brickwork still has cracks. Digital does not have cracks. There are no cracks betwen samples and there are no cracks between the lines in a FFT spectrum. |
#378
|
|||
|
|||
"Porky" wrote in message
When dealing with analog, I would agree with you wholeheartedly, but since any digital is only an approximation of the real thing, So is analog an approximation of the real thing. |
#379
|
|||
|
|||
"Bob Cain" wrote in message
Scott Dorsey wrote: So I think folks should continue investigating doppler distortion because it's an interesting problem even if not a terribly important one. Mostly agreed. The extent to which unbelievably small effects are claimed to be audible on the ProAudio mailing list, if they are other than imagination, does push any such effects like we are talking about into an arena at least worth discussing, if not important. The point is well taken. Gosh, I even brought it up about a week ago and our resident opamp wine tasters completely missed it. Go figure! It wouldn't take much _at all_ to swamp the things they consider very important. Speaker Doppler as insignificant as it is, is positively huge compared to the errors that a common nasty old 5532 or TL072 makes in most audio circuits. |
#380
|
|||
|
|||
Mark wrote: Bob, What makes you think Doppler does not occur in an infinite terminated tube. That the radiation impedence seen by the piston is identically that of the characteristic impedence of the gas. Consider the implcations of that. This is a similar case to a sliding RF transmission line. A radio wave in a transmission line that is changing length will experience Doppler shift. Well, yes if the reciever is in motion, other things begin to happen. I'm only, at this point, considering the situation where the transmitter and reciever are at a constant distance from each other. If the vernacular argument for Doppler distortion predicts the phenomenon for that case then it is incorrect, and it does. If you read my recent post to Randy you'll see that my thinking has changed signifigantly about Doppler distortion but not at all about the vernacular argument that started me questioning it. Based on the death of the "HF waves riding on LF ones" argument, which can be disproved by analysis of a piston in a tube, I threw out the whole idea of Doppler distortion but there is a mechanism that is a whole lot more complex than that which is so widely reprinted. I think you are stuck on the idea of linearity. Not any more. Something is clearly wrong with my belief that a system is linear if all complex exponentials are passed changed only by a complex multiplier. I'm certain that is an "only if" condition. I started to think about that last night but haven't come back to it. My provisional thought on it is that the accepted eigenfunction/eigenvalue definition applies to linear _time invariant_ systems and that we are somehow looking at a time varying system here (in terms of impedences.) Not sure that is right or not. Sorta shooting in the dark at this one so far. Anybody? What is it that allows a system to look linear when presented with sinusoids but not if presented with more complex signals? Please don't say "Doppler distortion". There has to be a more general systems principle involved. So in the speaker case, even if the cone and the air are perfectly linear loaded and the 2 waves add perfectly in amplitude, the Doppler that produces FM is still a non-linear process and produced new sidebands. Non-linear does not have to imply a defect of some kind. Perfect FM modulators are non-linear in the sense that the modulation process creates new sidebands. I agree, it is a mathematically defined constraint, not a defect. Now in a real speaker which also has amplitude non-linearities due to changing spider tension etc etc. it is defiantly a debatable issue as to weather or not the Doppler effect is SIGNIFICANT relative to all the other non-linear effects going on. I have no idea. Me either yet. I hope it is becoming clearer to anyone still with this discussion that I could give a rat's ass about being "right" on this, I just want to emerge correct. My ego isn't on the line, thank you very much, my understanding is. From a lot of the personal crap that has been slung around it is obvious that a lot of people don't understand the nature of scientific debate. This is it, folks. Not that there isn't a time and place for slinging crap in almost any scientific debate. :-) Thanks, Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#381
|
|||
|
|||
Porky wrote: If you want believable, try it for yourself and compare the results across a wide range of LF and HF tones. I suspect it's because we're dealing with digital waves, but there could be other factors involved. I suspect you are wrong on this one, Mike. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#382
|
|||
|
|||
"Bob Cain" wrote in message ... Mark Simonetti wrote: Of course it will. And its instanteious velocity is added to the instantaneous velocity due to the upper note. Okay, but it seems like people are saying that this isn't the case, and is the reason doppler distortion would NOT occur, which seems plain bizzare, IMHO. In some circumstances it does and in others it doesn't but that isn't the reason, at any rate, for either result. My position is that the speaker's motion does NOT directly produce the sound, so there is no tone riding on any other tone coming from the speaker. The complex motion of the speaker cone corresponds exactly with the compression and rarefaction of the air in front of it, and any instantaneous position of the cone is directly related to the corresponding level of compression or rarefaction of the complex acoustic wave. Adding a DC component doesn't matter, because the excursion levels will still be exactly the same as long as the linear excursion factor of the speaker is not exceeded, the center position will just be moved a bit. Slowly varying DC doesn't affect it either, since the variation becomes part of the acoustic wave being produced, whether we hear it or not. The flaw there was assuming that because we couldn't hear it, wasn't a sound wave. The only thing that can cause Doppler shift in a speaker is physically moving the speaker relative to the listener with some force other than that producing the complex wave driving the speaker motor. Here's another way of looking at it, putting a sound wave through a hole in the wall can't produce Doppler shift, no matter how many tones are in the waveform, and a speaker is effectively an artificial hole in the wall, in that the effective sound source isn't the speaker any more than it is the hole in the wall. Does anyone here think that if you stretched a thin diaphragm over a hole in a soundproof wall and had a band playing behind it, the diaphragm would cause Doppler Distortion? The speaker, provided it isn't exceeding its linear limits, is effectively exactly the same thing for all practical purposes. Instead of being driven by the sound source in the other room, it's driven by the electrical equivalent of the sound source in the other room. Can any of you provide an explanation of how an acoustic wave driving a diaphragm and passing the soundwave through it is in any way different than the diaphragm being driven by a motor being supplied with the exact electrical analog of that acoustic wave? And by that I mean that the difference will be such that the electrically driven one will produce Doppler distortion while the acoustically driven one doesn't. |
#383
|
|||
|
|||
"Randy Yates" wrote in message ... Bob Cain writes: Randy Yates wrote: Bob Cain writes: Randy Yates wrote: So it's dropped. Randy, if you are going to be so nasty, the least you could do is refute what I've said. Let me get something out of it anyway. How can I refute what you have refused to clearly restate? All right, I'll accept that you haven't read the thread. I've found a new way to state it anyway, that I think makes it clearer. I'm saying that the instantaneous velocity of the piston is transfered to the wave in the right position, since it is moving in step with it, to propegate that velocity out as an acoustic wave. It is _in_ the acoustic wave it is creating and it is at the right place at all times to impart the correct velocity _because_ it is in it. It doesn't matter what signals might have been mixed upstream to get the signal that controls the velocity of that piston. It will be moving in lock step with the wave defined by that signal and will always be in the right place to deliver the right velocity to the outgoing wave. If, on the other hand the piston is moving with a constant velocity superimposed on the signal velocity, it has no way to transfer that constant velocity to the air because contant velocity doesn't create a wave. At f=0 it runs out of punch. It ceases abruptly to transduce at all. In that case, the piston will always be in the wrong position to correctly impart the desired velocity signal and that error is Doppler shift. If the above is truly crazy, then I'll agree that I'm over the edge. But I don't think so. What, exactly, is wrong with it? I don't want to talk about trains and whistles. Refutations: a) Non-Causality: If the speaker cone and the wave are EXACTLY in sync, then the wave must instanteously "know" which way the piston is going. This is ludicrous. b) Discontinuous Frequency Offsets: Doppler operating as you describe would result in a sharp discontinuity in the perceived frequency as the relative acceleration changed from zero to non-zero. Similarly for the opposite case - i.e., while the acceleration is non-zero there would be no Doppler shift but as soon as it reaches zero the Doppler would discontinuously jump to the frequency predicted by the Doppler equation. This is ludicrous. c) Acoustic wave propagation in air is longitudinal and thus pressure-waves. The velocity of the transducer is not related to the velocity of the wave and imparts no information to the propagated signal. Gee, couldn't one to construe (c) to mean that Doppler distortion doesn't exist in a speaker, since it is a transducer? |
#384
|
|||
|
|||
"Arny Krueger" wrote in message ... "Porky" wrote in message When dealing with analog, I would agree with you wholeheartedly, but since any digital is only an approximation of the real thing, So is analog an approximation of the real thing. Nope, under perfect condidtions, analog is an exact duplication of the real thing, and that's something that cannot be said of digital, not matter how high the digital resolution, or how perfect the conditions, unless you can show me how if is possible to produce a digital signal with infinite bit width and infinitely high sampling rate. Wait, that would BE analog!!!:-) |
#385
|
|||
|
|||
"Bob Cain" wrote in message ... Porky wrote: If you want believable, try it for yourself and compare the results across a wide range of LF and HF tones. I suspect it's because we're dealing with digital waves, but there could be other factors involved. I suspect you are wrong on this one, Mike. I could very well be wrong, I'm basically just speculating on possible causes for the variations I see with different sampling frequencies.:-) |
#386
|
|||
|
|||
Bob Cain wrote: Randy, I really do appreciate the specificity of your argument. It's been really hard in this discussion to get ahold of anything with the generalities that have greeted everything I've tried to pin down. The understanding I'm arriving at would have come _much_ sooner with attack such as yours, correct or incorrect. It sounds like I'm saying that no one else ever did that and that's not what I meant at all. There was just an awful lot to sort through to get to the stuff that made me revise my thinking. I was just especially appreciative at the moment I wrote that. Too tired to be writing. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#387
|
|||
|
|||
Porky wrote: Then why would the analysis show differing results with identical waveforms whose only difference was that one was 16/44.1 and the other was 16/48? Damn good question. I dunno. Sure seems worth finding an answer to though. Note that the difference was considerable. Since all the wave simulations were done with Cool Edit, could it be something in the way Cool Edit generates the sine waves? Ya, know. At one point when I was playing with a possible model of the vernacular argument in Matlab I did get a nearly perfectly flat top. Not keeping notes, however, and not curious enough at that point because I was looking for something unrelated, I can't remember what expression I used. Dammit. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#388
|
|||
|
|||
Arny Krueger wrote: Speaker Doppler as insignificant as it is, is positively huge compared to the errors that a common nasty old 5532 or TL072 makes in most audio circuits. What's that say about the known and much larger non-linear effects in speakers that we all know and love? I think the guys that claim to hear these miniscule phase and dither differences played back through the best of the available speakers are blowing smoke... They argue about what they hear as a function of a couple of degrees of phase shift at Nyquist and the kind of dither applied to get to 24 bits! I notice that the one thing they are too totally polite to ever do is question each other's "golden" ears despite the differences among themselves in what they hear. Feet of clay all around, perhaps. It was my annoyance and disbelief in all of this that motivated me to look hard at Doppler distortion and find a way to quantify it. The rest, as they say, is history. :-) Actually, I am quite happy to find that it exists, if not for the usual reasons given, because of this original motivation. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#389
|
|||
|
|||
"Porky" wrote in message
"Arny Krueger" wrote in message ... "Porky" wrote in message When dealing with analog, I would agree with you wholeheartedly, but since any digital is only an approximation of the real thing, So is analog an approximation of the real thing. Nope, under perfect condidtions, analog is an exact duplication of the real thing, Irrelevant since it has never been observed, nor can it ever be observed in the real world. and that's something that cannot be said of digital, not matter how high the digital resolution, or how perfect the conditions, unless you can show me how if is possible to produce a digital signal with infinite bit width and infinitely high sampling rate. All real world analog signals have finite dynamic range and frequency response. Dynamic range in the analog domain is practically limited by thermal noise and the size of the largest signal that can be made with a good noise figure. In contrast, the dynamic range of a digital signal is limited only by our ability to implement and manipulate large numbers of bits. This just about unlimited. Wait, that would BE analog!!!:-) Wrong. In audio we know that flat frequency reponse over the audible range is more valuable and useful than rough or skewed frequency response from DC to light. We also know that good frequency response in the audible range is very hard to preserve in the analog domain as the processing, recording, and transmission steps build up. In contrast digital signals by default have perfectly flat response over any reasonble defined range of frequencies, right up into the microwave region, and down in to the frequency realm of siesmic events. |
#390
|
|||
|
|||
"Bob Cain" wrote in message
Arny Krueger wrote: Speaker Doppler as insignificant as it is, is positively huge compared to the errors that a common nasty old 5532 or TL072 makes in most audio circuits. What's that say about the known and much larger non-linear effects in speakers that we all know and love? They are there, and distract our ears from the itty-bitty ones. I think the guys that claim to hear these miniscule phase and dither differences played back through the best of the available speakers are blowing smoke... The rubber hits the road in a blind, level-matched, time-synched, bias-controlled listening test. They routinely lose out. They argue about what they hear as a function of a couple of degrees of phase shift at Nyquist and the kind of dither applied to get to 24 bits! Obviously they don't have a lot of good listening experiments under their belt. Cut your chops on good listening tests and you start singing a different song. I notice that the one thing they are too totally polite to ever do is question each other's "golden" ears despite the differences among themselves in what they hear. Feet of clay all around, perhaps. As long as they stay away from ABX and the like, their belief structures are preserved. It was my annoyance and disbelief in all of this that motivated me to look hard at Doppler distortion and find a way to quantify it. The rest, as they say, is history. :-) It's there! Actually, I am quite happy to find that it exists, if not for the usual reasons given, because of this original motivation. The point is that AM distortion, which dominates and is relatively large in speakers, and often pretty audible, is only present in good electronics in far smaller quantities. Masking rules, we hear distortion in speakers far more so than in non-clipping, non-noisy electronics. |
#391
|
|||
|
|||
Arny Krueger wrote:
Speaker Doppler as insignificant as it is, is positively huge compared to the errors that a common nasty old 5532 or TL072 makes in most audio circuits. Please excuse my ignorance, but what is a 5532 and a TL072 ? What problems are they causing ? I know I could look it up (and I just have) but others may be curious too. I find all this interesting. You'd be surprised (or perhaps not) how many people really aren't bothered about hearing nice clear audio through a decent system, never mind the (small?) distortions we are discussing in this NG. Sometimes I think its because people don't realise what they are missing. Even with my modest set up, I've bought a smile to the face of some people who've sat and listened to my relatively good system, compared to your average low end extremely coloured systems (complete with "3D" sound "enhancement" and all that).. I think others who refer to us, who do care, as anal about it perhaps do not appreciate music in the same way. To me there is nothing better than hearing the subtle beautiful tones of a well recorded well played piano or acoustic guitar; these are just plain invisible or distorted on a lot of todays cheap crap, and people are getting used to it. Disclaimer: IMHO, YMMV, etc. Cheers. -- Mark Simonetti. Freelance Software Engineer. |
#392
|
|||
|
|||
Mark Simonetti wrote:
I think others who refer to us, who do care, as anal about it perhaps do not appreciate music in the same way. To me there is nothing better than hearing the subtle beautiful tones of a well recorded well played piano or acoustic guitar; these are just plain invisible or distorted on a lot of todays cheap crap, and people are getting used to it. Disclaimer: IMHO, YMMV, etc. Cheers. Or maybe I'm just having a bad day and being cynical ;-) -- Mark Simonetti. Freelance Software Engineer. |
#393
|
|||
|
|||
"Mark Simonetti" wrote in message
Arny Krueger wrote: Speaker Doppler as insignificant as it is, is positively huge compared to the errors that a common nasty old 5532 or TL072 makes in most audio circuits. Please excuse my ignorance, but what is a 5532 and a TL072 ? Common, err mature audio op amp chips, still widely used particularly the former. What problems are they causing ? In correctly-designed circuits, that would be a good quesiton. I know I could look it up (and I just have) but others may be curious too. I find all this interesting. You'd be surprised (or perhaps not) how many people really aren't bothered about hearing nice clear audio through a decent system, never mind the (small?) distortions we are discussing in this NG. I'm sorry, but these sentence's structure and length makes me unclear about what it means. Sometimes I think its because people don't realise what they are missing. Are you suggesting that I'm missing something, or making these comments in a state of total ignorance? Even with my modest set up, I've bought a smile to the face of some people who've sat and listened to my relatively good system, compared to your average low end extremely coloured systems (complete with "3D" sound "enhancement" and all that).. That's not the scale of what we are talking about here. We're about a dozen cuts above at the baseline. I think others who refer to us, who do care, as anal about it perhaps do not appreciate music in the same way. I don't think you know who you are talking to, but frankly you should reserve such comments for your dogs, cats and other house pets. To me there is nothing better than hearing the subtle beautiful tones of a well recorded well played piano or acoustic guitar; these are just plain invisible or distorted on a lot of todays cheap crap, and people are getting used to it. Why do you presume that somehow this makes you different from me? Disclaimer: IMHO, YMMV, etc. A wise man checks out the lay of the land before he rushes spouting all sorts of self-serving insults. |
#394
|
|||
|
|||
"Arny Krueger" wrote in message ... "Porky" wrote in message "Arny Krueger" wrote in message ... "Porky" wrote in message When dealing with analog, I would agree with you wholeheartedly, but since any digital is only an approximation of the real thing, So is analog an approximation of the real thing. Nope, under perfect condidtions, analog is an exact duplication of the real thing, Irrelevant since it has never been observed, nor can it ever be observed in the real world. and that's something that cannot be said of digital, not matter how high the digital resolution, or how perfect the conditions, unless you can show me how if is possible to produce a digital signal with infinite bit width and infinitely high sampling rate. All real world analog signals have finite dynamic range and frequency response. Dynamic range in the analog domain is practically limited by thermal noise and the size of the largest signal that can be made with a good noise figure. In contrast, the dynamic range of a digital signal is limited only by our ability to implement and manipulate large numbers of bits. This just about unlimited. Wait, that would BE analog!!!:-) Wrong. In audio we know that flat frequency reponse over the audible range is more valuable and useful than rough or skewed frequency response from DC to light. We also know that good frequency response in the audible range is very hard to preserve in the analog domain as the processing, recording, and transmission steps build up. In contrast digital signals by default have perfectly flat response over any reasonble defined range of frequencies, right up into the microwave region, and down in to the frequency realm of siesmic events. Well, assuming that we can't reach infinity in digital bit depth, perfect analog is one bit depth better than perfect digital :-) |
#395
|
|||
|
|||
I think perhaps I didn't write that very well, you've midunderstood me,
sorry. I'm basically stating that in general people, as in the general public, do not appreciate quality audio, which is a sad state of affairs. Mark. -- Arny Krueger wrote: "Mark Simonetti" wrote in message Arny Krueger wrote: Speaker Doppler as insignificant as it is, is positively huge compared to the errors that a common nasty old 5532 or TL072 makes in most audio circuits. Please excuse my ignorance, but what is a 5532 and a TL072 ? Common, err mature audio op amp chips, still widely used particularly the former. What problems are they causing ? In correctly-designed circuits, that would be a good quesiton. I know I could look it up (and I just have) but others may be curious too. I find all this interesting. You'd be surprised (or perhaps not) how many people really aren't bothered about hearing nice clear audio through a decent system, never mind the (small?) distortions we are discussing in this NG. I'm sorry, but these sentence's structure and length makes me unclear about what it means. Sometimes I think its because people don't realise what they are missing. Are you suggesting that I'm missing something, or making these comments in a state of total ignorance? Even with my modest set up, I've bought a smile to the face of some people who've sat and listened to my relatively good system, compared to your average low end extremely coloured systems (complete with "3D" sound "enhancement" and all that).. That's not the scale of what we are talking about here. We're about a dozen cuts above at the baseline. I think others who refer to us, who do care, as anal about it perhaps do not appreciate music in the same way. I don't think you know who you are talking to, but frankly you should reserve such comments for your dogs, cats and other house pets. To me there is nothing better than hearing the subtle beautiful tones of a well recorded well played piano or acoustic guitar; these are just plain invisible or distorted on a lot of todays cheap crap, and people are getting used to it. Why do you presume that somehow this makes you different from me? Disclaimer: IMHO, YMMV, etc. A wise man checks out the lay of the land before he rushes spouting all sorts of self-serving insults. -- Mark Simonetti. Freelance Software Engineer. |
#396
|
|||
|
|||
"Mark Simonetti" wrote in message
I think perhaps I didn't write that very well, you've midunderstood me, sorry. You did more than express yourself poorly, you cast aspersons on other people. I'm basically stating that in general people, as in the general public, do not appreciate quality audio, which is a sad state of affairs. Why state the obvious in this context? |
#397
|
|||
|
|||
I think perhaps I didn't write that very well, you've midunderstood
me, sorry. You did more than express yourself poorly, you cast aspersons on other people. No, Mark, your point of view (though perhaps not geramaine to the discussion) was perfectly clear. Arny deliberately "misunderstood" you. He does this all the time. It's his snide way of attacking people he disagrees with, rather than directly confronting the issue. (He has a high IQ, but little insight.) You're not the only person who's been on the receiving end. |
#398
|
|||
|
|||
"Porky" writes:
"Randy Yates" wrote in message ... "Porky" writes: "Arny Krueger" wrote in message ... "Porky" wrote in message The experiment I suggested will give the results I gave, and if it is right at under the circumstances I suggested, it should be right under all circumstances with the same conditions, right? In other words, if it applies with a LF of .1 Hz or 1 Hz, it will still apply at LF's 20Hz or 50Hz, is that not correct? right. However, its a lot harder to properly measure doppler when the LF tone has a very low frequency. To measure it with a FFT you must use a FFT size that covers at least one cycle, and hopefully several cycles of the process. If the LF tone is 0.1 Hz, this means an absolute minimum of 10 seconds of data, and ideally 30 or more. At 44,100 Hz sampling, this would be a FFT composed of a minimum of 441,000 samples, and preferably several million samples. Consider the original example - the LF tone was 50 Hz. It had an 882 sample period. Note how much overkill there was when analyzed using a 65k sample FFT, or as I used a one million point FFT. One of the problems with FFT analysis that we've all overlooked is that we aren't really dealing with analog waveforms in our simulations, and we can get erroneous results when using high FFT numbers because we start playing in the digital "cracks", so to speak, The waveform being analog or digital makes no difference as long as sufficient bandwidth and dynamic range has been supplied by the A/D conversion. Rather, the problem you are ignorantly referring to is that an FFT implicitly assumes the input is periodic. If it isn't, you can get yourself befuddled. There is also the problem with using the FFT to estimate the spectrum of a random signal - it can be shown that there will be variance in the frequency estimates no matter how many points are used in the FFT (see, for example, "Signal Processing: Discrete Spectral Analysis, Detection, and Estimation," Mischa Schwartz and Leonard Shaw). -- Then why would the analysis show differing results with identical waveforms whose only difference was that one was 16/44.1 and the other was 16/48? Note that the difference was considerable. Hi Porky, Yes, I believe you. The reason is because, for the same number of input samples to the FFT (let's just use an example length of 1024, i.e., we're doing a 1024-point FFT), a different sample rate corresponds to a different period of time, and those 1024 points are considered periodic with period 1024*Ts (Ts = 1/Fs, Fs the sample rate). Here's a simple example. Let's say your sample rate is 1024 samples/second, and you're generating a 1 Hz sine wave. Then a 1024-point FFT will gather exactly one cycle of your 1 Hz sine wave and you will see a single, beautiful peak in bin 1 (counting starting with 0) of the FFT. That is because the 1024-point FFT implicitly assumes those 1024 samples are one period of a periodic waveform with period of 1024 * 1/Fs = 1024 * 1/1024 = 1 second. OK, now change the sample rate to 2048 samples/second. Now, 1024 samples corresponds to 1/2 cycle of your 1 kHz sine wave. When you perform the 1024-point FFT you will see energy all over the spectrum. Again, that is because the 1024-point FFT implicitly assumes those 1024 samples are one period of a periodic waveform with period of 1024 * 1/Fs = 1024 * 1/2048 = 0.5 second. You're thus seeing the spectrum of a half-sine (a rectified sine wave) at 1 Hz. Since all the wave simulations were done with Cool Edit, could it be something in the way Cool Edit generates the sine waves? As outlined above, it's not Cool Edit but rather the fundamental nature of the FFT. As a final note, we could replace "FFT" with "DFT" in all the above and the statements would still be true. The FFT is just an efficient way to compute the DFT. -- Randy Yates Sony Ericsson Mobile Communications Research Triangle Park, NC, USA , 919-472-1124 |
#399
|
|||
|
|||
"Mark Simonetti" wrote in message ... I think perhaps I didn't write that very well, you've midunderstood me, sorry. I'm basically stating that in general people, as in the general public, do not appreciate quality audio, which is a sad state of affairs. I agree wholeheartedly with that statement, however, what the traind and experienced discerning listener, and what the self-proclaimed "golden-eared audiophile" claim to hear are often totally different things. The difference between the two is that what the trained and experinced discerning listener claims to hear can be verified by scientific double-blind tests, and what the self-proclaimed "golden-eared audiophile" claims to hear often cannot. That isn't to say that all audiophiles are bogus, that isn't true, I know of audiophiles who are exceptionally accurate and discerning listeners. However the "audiophile" who buys $400 a foot speaker cable and claims to hear "a big difference" over good quality regular speaker cable, or the "audiophile" who claims to hear a phase difference of 2 degrees at far field in a live room, is full of it!:-) |
#400
|
|||
|
|||
Randy Yates writes:
[...] OK, now change the sample rate to 2048 samples/second. Now, 1024 samples corresponds to 1/2 cycle of your 1 kHz sine wave. Correction: 1 Hz sine wave, not 1 kHz. -- Randy Yates Sony Ericsson Mobile Communications Research Triangle Park, NC, USA , 919-472-1124 |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Stereophile Tries To Come Clean About The DiAural Fiasco | Audio Opinions | |||
Experimental Evidence for Dynamic Doppler Shift | Tech | |||
Bob Cain Is In Convulsions: A Doppler Piston Just Got Shoved Up His Tube | Tech | |||
Doppler Distoriton? | Tech | |||
Doppler Distortion - Fact or Fiction | Pro Audio |