Home |
Search |
Today's Posts |
#121
|
|||
|
|||
On Wed, 08 Jun 2005 11:31:23 GMT, "Tim Martin"
wrote: "Stewart Pinkerton" wrote in message .. . That's because you are ignoring reality, as in most of your posts. How else would you produce an analogue of the original sonic event? I explained that. One can measure the position of a vibrating diaphragm, Not with real-world equipment. These aren't mind games, we're talking about reality here. Besides, you'd have the same self-noise problems with your proposed measurement system, and *really* serious frequency response issues. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#122
|
|||
|
|||
On Wed, 08 Jun 2005 14:48:33 GMT, "Tim Martin"
wrote: "Richard Crowley" wrote in message ... Congratulations, you have re-invented the microphone. I think by "microphone" people normally mean a transducer that converts changes in air pressure into electrical signals. Indeed they do, although some microphones such as ribbons are more velocity-sensitive. It's possible to measure movement without that, and in fact there are commercial devices using optical measurement to measure such movements. Specify some. I'm well aware of such devices on machine tools, but not in this application. Such devices are not generally termed "microphones". They are, if they are used as sonic transducers. Basically, you have *no* idea what you're talking about - but that's been obvious for some time. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#123
|
|||
|
|||
Tim Martin wrote: "Arny Krueger" wrote in message ... What you don't seem to realize Tim is the fact that all of the limiations that you've been obsessing over relate to both analog and digital signals. That's incorrect: analog signals are continuous. And you keep making the same fundamental error over and over again, when you equate continuous with infinite resolution. It just ain't so, matter how many times you repeat it. Between any two points in time where the analog signal is changing, one can find another point in time where the value of the analog signal differs from its values at either of the first two points in time. And what you utterly fail or refuse to realize is that ALL the values in between those two points in a continuous representation are ENTIRELY defined by the bandwidth of the system. If I know the bandwidth, and I know the data on either side of of your two points, and those two points are separated by 1/2*bandwidth, then I can exactly predict the value at ANY arbitrary point in between. This is not true for digital representations of analog signals. Once again, other than stating it over and over and over agin, you have failed completely to deomstrate your assertion. Of course, there are limits in how accurately one can represent analog signals, whether the representation is digital or analog. Yup, and those limits are imposed first by the fact that NO signal can have infinite resolution to begin with, your irrational yet exuberant insistance on infinite energy and time not withstanding at all. Secondly, the limits are imposed by the bandwidth of the system and its dynamic range. But the limits are in communication channels and storage schemes, not in the signals themselves. Not if the signals themselves don't exist for infinite time, have infinite bandwidth and have inifite dynamic range. No matter how many times you repeat you incorrect assertion that "digital compresses", no matter how many times you misapply, deliberately or otherwise, the principles of Fourier decomposition, no matter how many times you invoke physically unrealizable conditions or theoretically ridiculous signals, no matter how many times you attempt to sidestep the issue with irrelevancies in an attempt to support your assertion, the assertion still remains incorrect. |
#124
|
|||
|
|||
Tim Martin wrote: However, if you think what I said is incorrect, please show it. In the case of the bird singing in the woods, demonstrate there are two points in time,where the analog signal in the interval between those two ponts in time is unchanging. No one ever said that the signal is unchanging. That's a strawman that you alone erected. What one CAN say, most definitely, is that the number of possible paths the signal CAN take between the two points is limited by the interval between those two points and the bandwidth of the signal. Make the bandwidth narrow enough, or make the two points close enough, and the signal is uniquely constrained. At that point, then, if you "sample" the signal at those intervals, you can have complete knowledge of the signal wihout having INFINITE knowledge of the signal. In other words, once you have sampled the signal at a rate fast enough as defined by the bandwidth of the signal, smapling any faster WILL NOT give you ANY more information about that signal. Any extra data is simply redundant and provides no new information that we could not have already derived from the data we have. Your repeated assertion is based on the fallacy that any signal, such as a bird singing in the woods, has infinite information content. It does not. Once you finally come to that realization, then you'll see the nonsense of your assertion. You do not need infinite data to represent finite information. Now, go ponder your position in the light that no signal has infinite information content. If you still hold to your position, please don't bother us until you are able to support that position with equal the rigor the likes of Shannon applied to the sampling theorem. |
#125
|
|||
|
|||
"Wessel Dirksen" wrote in...
Simpler yet, infinity is even ruled out by the very vehicle of sound propagation itself which can never be infinite. If anything is acoustic in nature, it can't be of infinite bandwidth otherwise it would violate Newton's basic law of conservation of energy. To say that any "sound" has infinite bandwidth is to say that inertia doesn't exist. Mr. Martin never directly addressed the question of whether he is posting his messages from the same universe the rest of us are in. He may not know what "inertia" is in his Martin Universe. :-) His responses and questions suggest that he may not operating from the same frame of reference as the rest of us. |
#126
|
|||
|
|||
"Wessel Dirksen" wrote in message oups.com... Simpler yet, infinity is even ruled out by the very vehicle of sound propagation itself which can never be infinite. If anything is acoustic in nature, it can't be of infinite bandwidth otherwise it would violate Newton's basic law of conservation of energy. To say that any "sound" has infinite bandwidth is to say that inertia doesn't exist. Sorry, I don't see why there being no lower limit to time resolution in an analog signal would violate conservation of energy. Tim |
#127
|
|||
|
|||
Tim Martin schreef: "Stewart Pinkerton" wrote in message ... That's incorrect: analog signals are continuous. Between any two points in time where the analog signal is changing, one can find another point in time where the value of the analog signal differs from its values at either of the first two points in time. Clearly, you have absolutely *no* understanding of the real physical world. Go learn about uncertainty, before you spout such nonsense here. As usual, instead of commenting on the point made, you put up a smokescreen. However, if you think what I said is incorrect, please show it. In the case of the bird singing in the woods, demonstrate there are two points in time,where the analog signal in the interval between those two ponts in time is unchanging. Tim Tim, I added this above but I'll repeat it because it may make this clear; aside from Nyquist, bits etc which I don't understand much of either. Any sound must be produced mechanically because something must push molecules into movement. Anything with mass cannot accellerate infinitely fast. You must know this; "objects at rest tend to stay at rest . . ." So it is impossible to have an infinite bandwidth soundwave purely because it could never be produced mechanically. A birds cloaca, or whatever it uses, has a HF limit where you can't accelerate its mass fast enough back and forth and thus the HF response has a very definite limit. On top of that, even the mass of molecules moving back and forth in a sound wave are limited in the same way. |
#128
|
|||
|
|||
wrote in message oups.com... What one CAN say, most definitely, is that the number of possible paths the signal CAN take between the two points is limited by the interval between those two points and the bandwidth of the signal. Make the bandwidth narrow enough, or make the two points close enough, and the signal is uniquely constrained. At that point, then, if you "sample" the signal at those intervals, you can have complete knowledge of the signal wihout having INFINITE knowledge of the signal. Great, a comment with some substance. OK, let's suppose our signal is a 1kHz sine wave, followed by a period of silence, followed by a 1kHz sine wave. The amplitude of the sine wave is the minimum that can be represented in our digital scheme. And let's suppose our sampling rate is 50000 times a second .... that is, one sample every 20 microseconds.. Now, I think that an analog signal with a silence of 200 microseconds will result in the same digital representation as an analog signal with a silence of 201 microseconds. Do you agree or not? Tim |
#129
|
|||
|
|||
Tim Martin wrote: wrote in message oups.com... What one CAN say, most definitely, is that the number of possible paths the signal CAN take between the two points is limited by the interval between those two points and the bandwidth of the signal. Make the bandwidth narrow enough, or make the two points close enough, and the signal is uniquely constrained. At that point, then, if you "sample" the signal at those intervals, you can have complete knowledge of the signal wihout having INFINITE knowledge of the signal. Great, a comment with some substance. OK, let's suppose our signal is a 1kHz sine wave, followed by a period of silence, followed by a 1kHz sine wave. The amplitude of the sine wave is the minimum that can be represented in our digital scheme. And let's suppose our sampling rate is 50000 times a second .... that is, one sample every 20 microseconds.. Now, I think that an analog signal with a silence of 200 microseconds will result in the same digital representation as an analog signal with a silence of 201 microseconds. Wrong, because the signal you describe CANNOT be represented in a system with finite bandwidth, a fundamental principle you abjectly refuse to grasp. More importantly, you've described a signal that you cannot even create in the real universe. Only unless you are willing to say that you have a signal produced in a "system" whose bandwidth is such (and such infinity), that signal cannot exist in ANY representation, continuous, discrete or otherwise. As soon as you come to the realization that the bandwidth MUST be limited in any realistic signal, you can then have a perfect discrete sampled representation of that signal. Your continued unsupportable protests to the contrary, e.g., what you think, is irrelevant. |
#130
|
|||
|
|||
"Tim Martin" wrote in message ... "Wessel Dirksen" wrote in message oups.com... Simpler yet, infinity is even ruled out by the very vehicle of sound propagation itself which can never be infinite. If anything is acoustic in nature, it can't be of infinite bandwidth otherwise it would violate Newton's basic law of conservation of energy. To say that any "sound" has infinite bandwidth is to say that inertia doesn't exist. Sorry, I don't see why there being no lower limit to time resolution in an analog signal would violate conservation of energy. A lower limit to time resolution is a consequence of the signal having finite duration. If the signal went on indefinitely it would have no lower limit. If the signal went on indefinitely and had finite energy in all or most finite time segments, then the total signal would have infinite amounts energy. Now if the presence of infinite amounts of energy doesn't in some sense violate the law of conservation of energy... ;-) |
#131
|
|||
|
|||
"Arny Krueger" wrote in message ... If the signal went on indefinitely and had finite energy in all or most finite time segments, then the total signal would have infinite amounts energy. Now if the presence of infinite amounts of energy doesn't in some sense violate the law of conservation of energy... ;-) Of course the duration cannot be infinite, ie. longer than the universe existence. MrT. |
#132
|
|||
|
|||
On Thu, 09 Jun 2005 11:35:28 GMT, "Tim Martin"
wrote: "Wessel Dirksen" wrote in message roups.com... Simpler yet, infinity is even ruled out by the very vehicle of sound propagation itself which can never be infinite. If anything is acoustic in nature, it can't be of infinite bandwidth otherwise it would violate Newton's basic law of conservation of energy. To say that any "sound" has infinite bandwidth is to say that inertia doesn't exist. Sorry, I don't see why there being no lower limit to time resolution in an analog signal would violate conservation of energy. That's because you don't understand the subject. No lower limit to time resolution = infinite bandwidth = infinite energy. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#133
|
|||
|
|||
wrote in message oups.com... Wrong, because the signal you describe CANNOT be represented in a system with finite bandwidth, a fundamental principle you abjectly refuse to grasp. More importantly, you've described a signal that you cannot even create in the real universe. So, you're saying I cannot create a signal comprising a 1kHz sine wave with small amplitude, followed by a period of silence, followed by a 1kHz sine wave with small amplitude. Tim |
#134
|
|||
|
|||
Tim Martin wrote: wrote in message oups.com... Wrong, because the signal you describe CANNOT be represented in a system with finite bandwidth, a fundamental principle you abjectly refuse to grasp. More importantly, you've described a signal that you cannot even create in the real universe. So, you're saying I cannot create a signal comprising a 1kHz sine wave with small amplitude, followed by a period of silence, followed by a 1kHz sine wave with small amplitude. No, you're saying that you can. So, prove us wrong. Produce for us this signal EXACTLY as you have described it (ignoring the fact, for the moment, that your description of such signal is QUITE inexact) through any means at your disposal. By signal, I don't mean an equation (continuous or otherwise), or a drawing on a piece of paper or any such dodge. I mean a signal. It can be acoustical, electrical, mechanical, what have you. By the way, as you HAVE provided a woefully incomplete description, how about we better define the signal? How long is the 1 kHz signal? Exactly when does it end? How long is the silence? Exactly when does it end? How silent is silent? How long is the second period of the 1 kHz since wave? How small is its amplitude? Fine, you say it can, so produce it. Let us know. |
#135
|
|||
|
|||
On Thu, 09 Jun 2005 16:08:47 GMT, "Tim Martin"
wrote: wrote in message roups.com... Wrong, because the signal you describe CANNOT be represented in a system with finite bandwidth, a fundamental principle you abjectly refuse to grasp. More importantly, you've described a signal that you cannot even create in the real universe. So, you're saying I cannot create a signal comprising a 1kHz sine wave with small amplitude, followed by a period of silence, followed by a 1kHz sine wave with small amplitude. If you allow the sine wave a finite period of attack and decay, and the 'silence' to have a noise floor, then of course you can. Otherwise, no one can. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#136
|
|||
|
|||
On Thu, 09 Jun 2005 11:48:47 GMT, "Tim Martin"
wrote: OK, let's suppose our signal is a 1kHz sine wave, followed by a period of silence, followed by a 1kHz sine wave. The amplitude of the sine wave is the minimum that can be represented in our digital scheme. And let's suppose our sampling rate is 50000 times a second .... that is, one sample every 20 microseconds.. Now, I think that an analog signal with a silence of 200 microseconds will result in the same digital representation as an analog signal with a silence of 201 microseconds. Do you agree or not? No. You're describing an analogue waveform ending towards infinite bandwidth. It can't exist in analogue or digital represention (If it COULD, you'd be thanking the digital system for ironing out such glitches, not complaining about low resolution :-) |
#137
|
|||
|
|||
On Thu, 09 Jun 2005 16:08:47 GMT, "Tim Martin"
wrote: So, you're saying I cannot create a signal comprising a 1kHz sine wave with small amplitude, followed by a period of silence, followed by a 1kHz sine wave with small amplitude. Not if you want that period of silence to be as short as we've been discussing. No you can't. You can describe it as a theoretical possibility. But, lacking resources tending to the infinite, you can't create it. |
#138
|
|||
|
|||
"calmar" wrote in message
... Hi, I'm wondering, how a good computer with a good graphic card + good speakers can compare to a good HiFi System? Since good soundcards can be quite expensive, and that only for the card itself, I would suspect, that that good computer/soundcard and good speaker combo can be as good as a good HiFi System? So I really don't know much about these things. Thanks for any hints. calmar For a long time now I have plugged my various PC sound card outputs (line level, not speaker output) into the AUX input of "surplus" stereo receivers (1970's to 80's vintage, around 15 to 25 wpc) and used whatever "spare" stereo speakers that were around at the time. The results have been generally good for CD's played on the PC. But all this stuff is in in the office so I can't do an A-B comparison with my main system. Anyway, try it. What can you lose? Cheers, Roger |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Question FAQ: rec.audio.* Recording 2/99 (part 7 of 13) | Pro Audio | |||
HiFi - BUY - SELL - EXCHANGE - Surrey, UK - Linn, Naim, Quad, Rega, SME, Neat etc... | Marketplace | |||
FA: Complete Car Computer - MP3, WMA, GPS, Auto PC | Car Audio | |||
FA: Complete Car Computer - MP3, WMA, GPS, Auto PC | Marketplace | |||
External RCA leads from Computer to HiFi | General |