Home |
Search |
Today's Posts |
#321
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Bill,
That was just a cheap shot on my part, as in jealous. ROFL. No offense was taken. When I think of "cheap shots" I think of guys who hide behind anonymous names and insult those who know more than they do, and say only "you're wrong" without saying why. You're not like that at all! :-) --Ethan |
#322
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Arny Krueger wrote:
"bill" wrote in message om... Arny Krueger wrote: "bill" wrote in message . com... Arny Krueger wrote: "bill" wrote in message .. . It may be personal choice, but I hate the square clipping of overdriven transistor amps. I also hate the clipping, square and otherwise, of overdriven digital audio gear. My solution to all of the above is to simply not overdrive amps or digital gear. The dynamic range of either is at least 10 dB better than the traditional alternatives, so avoiding clipping isn't all that hard to do. Yes and no. You can listen to something at 200 WRMS and then something happens like a big drum and rails the amp, even if only for one note if it has a KW peak component. I don't think you understand what I did say and what I didn't say. I didn't say that clipping is always avoided, I said that since transistor amps generally have at least 10 dB more dynamic range than tubed amps, its easier to avoid clipping with SS amps. I said that since digital recorders generally have 10 dB more dynamic range than analog recorders, it is easier to avoid clipping with digital recorders. I understood perfectly what I said. What can happen, and this is why bi-amps, and even tri-amps were used by some serious audiophiles is that a large low frequency component may be near the peak of the amplifier range when along comes a big mid frequency component, and then on top of that maybe somebody clashes some symbols (spell checker buys this) and then rails the single amp that should have had overhead. Dynamic range doesn't mean much if an amplifier is rated at 200 WRMS then that is pretty much it. Anything, even peaks, above that will give distortion, just soft or hard. Bad design is bad design. With tubes, the world of power amps pretty well ended at 300 pwc, and amps that big were seriously rare even in the days of tubes. With SS the current power output ceiling is around 10 killowatts, with plentiful supply. Tell that to some of my now nearly deaf friends who danced in discos in front of KW class band amps with TUBES. Super expensive but still tubes back in the early 70's. With SS I can build a 50 KW PWM amp if I want since it can be done easily enough, except for finding a power plug in. Football stadiums probably use even more than that, but low fidelity. Real world example - in the days of tubes, a good component stereo home hifi had 25 wpc x 1 or 2 channels for a total of 50 watts or less. Today, comparable class home hifis have 100 wpc X 5 for a total of 500 watts, and that doesn't include the power amp for the subwoofer. BTW, some will say that home hifis had more like 30 or 35 wpc, but if tests typical old tubed amps to modern standards, namely 20-20 KHz at rated power at 0.1 % distortion, they turn into more like 25 wpc amps. I was spoiled. I had two separate Ampex pro amps that would heat up the room and blow your eardrums, not a home unit which as you say would be lucky to get 25W. The only Ampex pro amps I recall or can find record of were the 620 powered speakers - which were what, 12 watts? I had some in 1976 that were bare chassis, maybe 50 Watts, and were totally bulletproof. At one point I drove a high voltage transformer with them to drive two 4 foot fluorescent light tubes for a stereo light show and the amps didn't care. A solid state unit would have just self destructed and the dancing light patterns made it all worthwhile. Too long ago to remember the exact model. In the days of analog tape, 80 dB dynamic range (without compression) would be truely outstanding performance for analog tape. Today, even consumer audio interfaces have 90 dB dynamic range, and deliver that 90 dB dynamic range over the fully range of audio frequencies with low distortion, which analog tape can't possibly do. Tape had a certain noise floor determined by hiss but the new digital uses sampling techniques with ADC's that are not necessarily linear like the instrument ADC's I used in my work. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. 16 bits is about the limit on true accuracy. No, 19 or 20 bits is pretty common in higher end audio interfaces. But, we're still talking 4 channels in and out for under $1K. The DC accuracy stinks, which I know is not a concern, but check out the INL and DNL figures. Also, regardless of what Nyquist says you better sample a 20KHz signal at a bit more than twice the frequency, more like 4 o 8 times as much, Horsefeathers. Do you believe everything you were taught in school and never question the fact that Nyquist applies nicely to a sine wave and not that well to music??? which puts you into the 196 KHz range and then you do not get 24 bits accuracy like some would have you believe. Actual unweighted 24 bit dynamic range is very elusive. I know of no regular commercial audio production equipment that has even 23 bit dynamic range, without resetting some gain control someplace. I know of no power amps with even 21 bit dynamic range. Mics and rooms run out of gas well below 16 bits. True. That is why I would run an instrumentation amp at a 16 bit sample and 196 KHz or so for a master. The 24 bit stuff is just marketing BS. You can get 24 bits of resolution with only 16 (or less) bits of true accuracy. Agreed that there is a lot of equipment out there that makes very noisy 24 bit data words. Same useful info could fit into 16 in many cases. Exactly what I was trying to get to in the first place. Screw the marketing people and use common sense. -- Bill (Sleepless biker) Baka |
#323
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Ethan Winer wrote:
Bill, That was just a cheap shot on my part, as in jealous. ROFL. No offense was taken. When I think of "cheap shots" I think of guys who hide behind anonymous names and insult those who know more than they do, and say only "you're wrong" without saying why. You're not like that at all! :-) --Ethan I have at least worked in the industry, even if it was over 15 years ago. -- Bill (Sleepless biker) Baka |
#324
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Carey Carlan wrote:
"William Sommerwerck" wrote in : I know of very few people who do not do experiments without first having a hypothesis in mind. And those that don't are the better scientists. You don't know what science is, do you? Hmm, I'm thinking that Bill is heading us into post-modern *wisdom*. You know, Arny, for all the interchanges we've had over the years, it's obvious you haven't paid the least attention to anything I've said. You're a poor thinker and an even worse "scientist". You lack insight and imagination. Science is not defined by the scientific method. The two greatest scientists of all time performed very few experiments. Their genius lay in intuition -- seeing things that other people ignored or didn't see at all. Intuition is one of the great keys of progress. It takes leaps beyond the practical to see things like gravity and relativity. Great minds like Gallileo, Newton and Einstein grasped concepts beyond their fellows by leaps of intuition. My contention, and I think Arny will agree, is that intuition without experimental proof is philosophy. It doesn't become science until mathematics and empirical evidence support the theory. Until then, you're in the realm of Aristotle with the four elemental substances and smoke rising because that's its "natural place". Exactamundo. -- Aaron |
#325
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
bill "just use comcast", "you know the drill for no spam bots." wrote:
Arny Krueger wrote: The only Ampex pro amps I recall or can find record of were the 620 powered speakers - which were what, 12 watts? I had some in 1976 that were bare chassis, maybe 50 Watts, and were totally bulletproof. At one point I drove a high voltage transformer with them to drive two 4 foot fluorescent light tubes for a stereo light show and the amps didn't care. A solid state unit would have just self destructed and the dancing light patterns made it all worthwhile. Too long ago to remember the exact model. Ampex made a lot of amps for theatre sound systems back in the sixties, including all of the hardware for Vista-Vision and complete turnkey Cinemascope sound systems. They made some big power amps and some little ones but they mostly sold them as part of turnkey assemblies. I scrapped a bunch of them for the power transformers back then. Wish I had kept them. --scott -- "C'est un Nagra. C'est suisse, et tres, tres precis." |
#326
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"Laurence Payne" lpayne1NOSPAM@dslDOTpipexDOTcom wrote in message
... On Wed, 25 Oct 2006 06:51:09 GMT, "Paul Stamler" wrote: Other experiments are in the nature of fishing expeditions, e.g.: "We've extracted a new antibiotic from a mold found on a rotten canteloupe; we're now going to drop it into 84 petri dishes, each containing a different bacterial culture, and see if it inhibits the growth of any of them." There's a hypothesis implicit there. Implicit but less explicit than "This antibiotic will work on shigellosis." More like "Maybe this antibiotic will work on something." Yes, still a hypothesis, but a much more open-ended one, which makes it a fishing expedition. Look, *any* scientific experiment contains the implicit hypothesis "There is something to be learned here." That's a given. It's the explicit hypotheses that differentiate focused experiments from fishing. Peace, Paul |
#327
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
bill wrote:
Arny Krueger wrote: ....snip.. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. In the limit (per Nyquist) you will not be able to determine the amplitude of that 20k signal with said 40k sample rate. You need to be a little below the Nyquist limit to gain that information, how far below determines how fast you can acquire the amplitude knowledge. Put another way, from the view point of bandwidth limiting, all modulation products need to be below the Nyquist limit else (any out of band) information will be lost. Later... Ron Capik -- |
#328
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"bill" wrote in message m... Arny Krueger wrote: "bill" wrote in message om... Arny Krueger wrote: "bill" wrote in message . com... Arny Krueger wrote: "bill" wrote in message .. . In the days of analog tape, 80 dB dynamic range (without compression) would be truely outstanding performance for analog tape. Today, even consumer audio interfaces have 90 dB dynamic range, and deliver that 90 dB dynamic range over the fully range of audio frequencies with low distortion, which analog tape can't possibly do. Tape had a certain noise floor determined by hiss but the new digital uses sampling techniques with ADC's that are not necessarily linear like the instrument ADC's I used in my work. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. Not true, if the obvious bandwidth restriction and its consequences in the frequency domain are fully considered. More below. 16 bits is about the limit on true accuracy. No, 19 or 20 bits is pretty common in higher end audio interfaces. But, we're still talking 4 channels in and out for under $1K. The DC accuracy stinks, which I know is not a concern, but check out the INL and DNL figures. Not much of a problem because audio DACs are pretty much all built on the Sigma-Delta model, which inherently has no missing codes and has an inherently monotonic transfer function. It's not practical or reasonable to use the Sigma-Delta model for many instrumention, video and RF applications. Also, regardless of what Nyquist says you better sample a 20KHz signal at a bit more than twice the frequency, more like 4 o 8 times as much, Horsefeathers. Do you believe everything you were taught in school and never question the fact that Nyquist applies nicely to a sine wave and not that well to music??? There are several ironies here, one being that digital audio was in its infancy when I went to school, and was not taught to undergraduates. Another is that I've spent about 10 years studying the technical and subjective implications of the application of digital technology to audio in a number of very hands-on ways. So the answer is that I was never origionally taught anything at all about digital audio in school because for all practical purposes there wasn't much of it at all at the time, but I've done a fair job of make-up. See former comments about intensive tests of like 100 different pieces of audio gear, mostly digital. One can directly study the impact of sample rates and therefore Nyquist frequencies on audio by downloading and listening to files from this web page: http://64.41.69.21/technical/sample_rates/index.htm which puts you into the 196 KHz range and then you do not get 24 bits accuracy like some would have you believe. Actual unweighted 24 bit dynamic range is very elusive. I know of no regular commercial audio production equipment that has even 23 bit dynamic range, without resetting some gain control someplace. I know of no power amps with even 21 bit dynamic range. Mics and rooms run out of gas well below 16 bits. True. That is why I would run an instrumentation amp at a 16 bit sample and 196 KHz or so for a master. The 24 bit stuff is just marketing BS. This is an example of the performance 24/192 audio device that is within a bit or two of the SOTA: http://www.pcavtech.com/soundcards/lynxtwo/ You can get 24 bits of resolution with only 16 (or less) bits of true accuracy. Agreed that there is a lot of equipment out there that makes very noisy 24 bit data words. Same useful info could fit into 16 in many cases. Exactly what I was trying to get to in the first place. Screw the marketing people and use common sense. Agreed. |
#329
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"Scott Dorsey" wrote in message ... bill "just use comcast", "you know the drill for no spam bots." wrote: Arny Krueger wrote: The only Ampex pro amps I recall or can find record of were the 620 powered speakers - which were what, 12 watts? I had some in 1976 that were bare chassis, maybe 50 Watts, and were totally bulletproof. At one point I drove a high voltage transformer with them to drive two 4 foot fluorescent light tubes for a stereo light show and the amps didn't care. These must be the smaller Ampex amps I've now found documented in a number of places. Too bad you don't remember more about them (see below). A solid state unit would have just self destructed and the dancing light patterns made it all worthwhile. By 1976 SS amps were getting to be fairly durable. Ampex made a lot of amps for theatre sound systems back in the sixties, including all of the hardware for Vista-Vision and complete turnkey Cinemascope sound systems. They made some big power amps and some little ones but they mostly sold them as part of turnkey assemblies. The *big ones* are documented in a number of places including: Multichannel Sound Reproduction, JAES Volume 2 Number 1 pp. 20-24; January 1954, Author: Selsted, Walter T They are pictured in the article from the bottom, and are said by various sources to be 100-120 wpc, based on 4x 807. Sounds about right. I scrapped a bunch of them for the power transformers back then. Wish I had kept them. No value for the OPTs at the time? |
#330
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Scott Dorsey wrote:
bill "just use comcast", "you know the drill for no spam bots." wrote: Arny Krueger wrote: The only Ampex pro amps I recall or can find record of were the 620 powered speakers - which were what, 12 watts? I had some in 1976 that were bare chassis, maybe 50 Watts, and were totally bulletproof. At one point I drove a high voltage transformer with them to drive two 4 foot fluorescent light tubes for a stereo light show and the amps didn't care. A solid state unit would have just self destructed and the dancing light patterns made it all worthwhile. Too long ago to remember the exact model. Ampex made a lot of amps for theatre sound systems back in the sixties, including all of the hardware for Vista-Vision and complete turnkey Cinemascope sound systems. They made some big power amps and some little ones but they mostly sold them as part of turnkey assemblies. I scrapped a bunch of them for the power transformers back then. Wish I had kept them. --scott Heh, Me too since they were 'unkillable'. -- Bill (Sleepless biker) Baka |
#331
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"Carey Carlan" wrote in message ... "William Sommerwerck" wrote in : Intuition is one of the great keys of progress. It takes leaps beyond the practical to see things like gravity and relativity. Great minds like Gallileo, Newton and Einstein grasped concepts beyond their fellows by leaps of intuition. My contention, and I think Arny will agree, is that intuition without experimental proof is philosophy. I'm not so sure that philosoply is that lame in every case. It doesn't become science until It doesn't become a provisional fact until... mathematics and empirical evidence support the theory. Especially the empirical evidence part. IME the relationships between math models and empirical evidence can be less than perfect. Who said it didn't? You said "Science is not defined by the scientific method." The only difference between philosophy and science is the execution of the scientific method. Seems a bit over-general, but I agree with the basic concept. |
#332
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Ron Capik wrote:
bill wrote: Arny Krueger wrote: ....snip.. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. In the limit (per Nyquist) you will not be able to determine the amplitude of that 20k signal with said 40k sample rate. You need to be a little below the Nyquist limit to gain that information, how far below determines how fast you can acquire the amplitude knowledge. Put another way, from the view point of bandwidth limiting, all modulation products need to be below the Nyquist limit else (any out of band) information will be lost. Later... Ron Capik -- The OBVIOUS problem is that if you only have exactly 2 samples per 20KHz sine wave you might sample on the zero crossings and not on the peaks, which would give you an output of 0.0 and not the true amplitude. Simple common sense says that you need to bag at least 4 samples to give a good chance of at least some accuracy. Math can be if not wrong, not optimal in the real world. I live in the real world. -- Bill (Sleepless biker) Baka |
#333
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"Ron Capik" wrote in message ... bill wrote: Arny Krueger wrote: ....snip.. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. In the limit (per Nyquist) you will not be able to determine the amplitude of that 20k signal with said 40k sample rate. You need to be a little below the Nyquist limit to gain that information, how far below determines how fast you can acquire the amplitude knowledge. Put another way, from the view point of bandwidth limiting, all modulation products need to be below the Nyquist limit else (any out of band) information will be lost. Agreed. IOW you need to handle the sidebands created by modulation if the sine wave is modulated. The brick wall filter can be capable of making AM into SSB! Music can be thought of as a collection of modulated sine waves, but most of the time the modulations are at a relatively low rate, 0.2 x the carrier. IOW, most musical sounds include more than 5 cycles of the fundamental. Most things that generate tones at say 16 KHz or greater are very resonant, and can't generate sharply-defined tone bursts. The ear, being based on a collection of resonant hairs, doesn't accurately follow them such as they are. The ear can't send info at a high enough data rate to the brain, even if it could follow sharply-defined tone bursts. The ear is losing track of accurate timing information by 1 KHz, and falls-off rapidly above that. When all of this gets critical, the carrier frequencies are so high that the ear doesn't hear a lot, and not very accurately. |
#334
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"bill" wrote in message t... Ron Capik wrote: bill wrote: Arny Krueger wrote: ....snip.. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. In the limit (per Nyquist) you will not be able to determine the amplitude of that 20k signal with said 40k sample rate. You need to be a little below the Nyquist limit to gain that information, how far below determines how fast you can acquire the amplitude knowledge. Put another way, from the view point of bandwidth limiting, all modulation products need to be below the Nyquist limit else (any out of band) information will be lost. Later... Ron Capik -- The OBVIOUS problem is that if you only have exactly 2 samples per 20KHz sine wave you might sample on the zero crossings and not on the peaks, which would give you an output of 0.0 and not the true amplitude. Simple common sense says that you need to bag at least 4 samples to give a good chance of at least some accuracy. Math can be if not wrong, not optimal in the real world. I live in the real world. Simple common sense overstates the problem by a factor of slightly more than 2. IOW, you can get everything you need to know about phase and amplitude by slightly more than 2 samples per wave. However, if the wave is modulated and its fundamental is too close to the corner frequency of the brickwall filter, then you may lose some information about the modulation. |
#335
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Arny Krueger wrote:
"bill" wrote in message m... Arny Krueger wrote: "bill" wrote in message om... Arny Krueger wrote: "bill" wrote in message . com... Arny Krueger wrote: "bill" wrote in message .. . In the days of analog tape, 80 dB dynamic range (without compression) would be truely outstanding performance for analog tape. Today, even consumer audio interfaces have 90 dB dynamic range, and deliver that 90 dB dynamic range over the fully range of audio frequencies with low distortion, which analog tape can't possibly do. Tape had a certain noise floor determined by hiss but the new digital uses sampling techniques with ADC's that are not necessarily linear like the instrument ADC's I used in my work. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. Not true, if the obvious bandwidth restriction and its consequences in the frequency domain are fully considered. More below. As I just mentioned in another post on this what happens if you sample your 20KHz sine wave with a 40KHz sample rate that just happens to sample the zero crossings and not the peaks??? I think you know. 16 bits is about the limit on true accuracy. No, 19 or 20 bits is pretty common in higher end audio interfaces. But, we're still talking 4 channels in and out for under $1K. The DC accuracy stinks, which I know is not a concern, but check out the INL and DNL figures. Not much of a problem because audio DACs are pretty much all built on the Sigma-Delta model, which inherently has no missing codes and has an inherently monotonic transfer function. It's not practical or reasonable to use the Sigma-Delta model for many instrumention, video and RF applications. Which has been a bummer for me since absolute accuracy is what I need for my instrumentation design work. Low frequency, under 500Hz but it has to be damned accurate. Also, regardless of what Nyquist says you better sample a 20KHz signal at a bit more than twice the frequency, more like 4 o 8 times as much, Horsefeathers. Do you believe everything you were taught in school and never question the fact that Nyquist applies nicely to a sine wave and not that well to music??? There are several ironies here, one being that digital audio was in its infancy when I went to school, and was not taught to undergraduates. Another is that I've spent about 10 years studying the technical and subjective implications of the application of digital technology to audio in a number of very hands-on ways. So the answer is that I was never origionally taught anything at all about digital audio in school because for all practical purposes there wasn't much of it at all at the time, but I've done a fair job of make-up. See former comments about intensive tests of like 100 different pieces of audio gear, mostly digital. One can directly study the impact of sample rates and therefore Nyquist frequencies on audio by downloading and listening to files from this web page: http://64.41.69.21/technical/sample_rates/index.htm Yes and no. Your ears are not really going to pick up much distortion over about 5 KHz anyway, just whether it is there or not. I can damn well hear distortion at 1 KHz, but not above 5 KHz even though my hearing goes to about 20 KHz, still, since I avoided disco's when they were the hot ticket. People may claim to be able to hear distortion at high frequency, but I doubt it, and only a spectrum analyzer will tell the truth. which puts you into the 196 KHz range and then you do not get 24 bits accuracy like some would have you believe. Actual unweighted 24 bit dynamic range is very elusive. I know of no regular commercial audio production equipment that has even 23 bit dynamic range, without resetting some gain control someplace. I know of no power amps with even 21 bit dynamic range. Mics and rooms run out of gas well below 16 bits. True. That is why I would run an instrumentation amp at a 16 bit sample and 196 KHz or so for a master. The 24 bit stuff is just marketing BS. This is an example of the performance 24/192 audio device that is within a bit or two of the SOTA: http://www.pcavtech.com/soundcards/lynxtwo/ I just took a quick look, long site page, and noticed right off the top that the 44KHz card beat the other 2 that sampled at a higher rate, but that was at 1KHz and 20Hz. The IM distortion I have to wonder about. The 11.025 KHz showed some side spurs which I would expect, but that is the nature of digital. Gain some things lose some others. Tradeoffs. You can get 24 bits of resolution with only 16 (or less) bits of true accuracy. Agreed that there is a lot of equipment out there that makes very noisy 24 bit data words. Same useful info could fit into 16 in many cases. Exactly what I was trying to get to in the first place. Screw the marketing people and use common sense. Agreed. -- Bill (Sleepless biker) Baka |
#336
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"bill" wrote in message t... Arny Krueger wrote: "bill" wrote in message m... Arny Krueger wrote: "bill" wrote in message om... Arny Krueger wrote: "bill" wrote in message . com... Arny Krueger wrote: "bill" wrote in message .. . In the days of analog tape, 80 dB dynamic range (without compression) would be truely outstanding performance for analog tape. Today, even consumer audio interfaces have 90 dB dynamic range, and deliver that 90 dB dynamic range over the fully range of audio frequencies with low distortion, which analog tape can't possibly do. Tape had a certain noise floor determined by hiss but the new digital uses sampling techniques with ADC's that are not necessarily linear like the instrument ADC's I used in my work. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. Not true, if the obvious bandwidth restriction and its consequences in the frequency domain are fully considered. More below. As I just mentioned in another post on this what happens if you sample your 20KHz sine wave with a 40KHz sample rate that just happens to sample the zero crossings and not the peaks??? I think you know. I know and also know that any lower frequency than 20 KHz can be fully sampled for amplitude and phase. Your example is one delta (where delta is the smallest non-zero number imagninable) from being false. 16 bits is about the limit on true accuracy. No, 19 or 20 bits is pretty common in higher end audio interfaces. But, we're still talking 4 channels in and out for under $1K. The DC accuracy stinks, which I know is not a concern, but check out the INL and DNL figures. Not much of a problem because audio DACs are pretty much all built on the Sigma-Delta model, which inherently has no missing codes and has an inherently monotonic transfer function. It's not practical or reasonable to use the Sigma-Delta model for many instrumention, video and RF applications. Which has been a bummer for me since absolute accuracy is what I need for my instrumentation design work. Low frequency, under 500Hz but it has to be damned accurate. AFAIK Sigma-Delta has no problems with accuracy at low frequencies. It just has implementation problems at very high frequencies. |
#337
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Arny Krueger wrote:
"Scott Dorsey" wrote in message ... bill "just use comcast", "you know the drill for no spam bots." wrote: Arny Krueger wrote: The only Ampex pro amps I recall or can find record of were the 620 powered speakers - which were what, 12 watts? I had some in 1976 that were bare chassis, maybe 50 Watts, and were totally bulletproof. At one point I drove a high voltage transformer with them to drive two 4 foot fluorescent light tubes for a stereo light show and the amps didn't care. These must be the smaller Ampex amps I've now found documented in a number of places. Too bad you don't remember more about them (see below). I'm not sure what they were, but they were open chassis about a foot square with about four big tubes each and some big output transformers. A solid state unit would have just self destructed and the dancing light patterns made it all worthwhile. By 1976 SS amps were getting to be fairly durable. Not that durable. I killed an $800 MacIntosh with that trick and it cost me a fair amount to replace the output transistors. The problem was using the transformer to drive the fluorescent lights which basically went up to a voltage and then arced down to a lower voltage to make light. Ampex made a lot of amps for theatre sound systems back in the sixties, including all of the hardware for Vista-Vision and complete turnkey Cinemascope sound systems. They made some big power amps and some little ones but they mostly sold them as part of turnkey assemblies. The *big ones* are documented in a number of places including: Multichannel Sound Reproduction, JAES Volume 2 Number 1 pp. 20-24; January 1954, Author: Selsted, Walter T They are pictured in the article from the bottom, and are said by various sources to be 100-120 wpc, based on 4x 807. Sounds about right. Those may be what I had. Bob Pease of National Semiconductor likes those amps too, for the unbreakable factor. He writes a column in Electronic Design. I scrapped a bunch of them for the power transformers back then. Wish I had kept them. No value for the OPTs at the time? There sure is now if you look on Ebay. -- Bill (Sleepless biker) Baka |
#338
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Arny Krueger wrote:
"Ron Capik" wrote in message ... bill wrote: Arny Krueger wrote: ....snip.. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. In the limit (per Nyquist) you will not be able to determine the amplitude of that 20k signal with said 40k sample rate. You need to be a little below the Nyquist limit to gain that information, how far below determines how fast you can acquire the amplitude knowledge. Put another way, from the view point of bandwidth limiting, all modulation products need to be below the Nyquist limit else (any out of band) information will be lost. Agreed. IOW you need to handle the sidebands created by modulation if the sine wave is modulated. The brick wall filter can be capable of making AM into SSB! Music can be thought of as a collection of modulated sine waves, but most of the time the modulations are at a relatively low rate, 0.2 x the carrier. IOW, most musical sounds include more than 5 cycles of the fundamental. Most things that generate tones at say 16 KHz or greater are very resonant, and can't generate sharply-defined tone bursts. The ear, being based on a collection of resonant hairs, doesn't accurately follow them such as they are. The ear can't send info at a high enough data rate to the brain, even if it could follow sharply-defined tone bursts. The ear is losing track of accurate timing information by 1 KHz, and falls-off rapidly above that. When all of this gets critical, the carrier frequencies are so high that the ear doesn't hear a lot, and not very accurately. Agreed. I can hear it, but not any distortion or much direction information. -- Bill (Sleepless biker) Baka |
#339
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Arny Krueger wrote:
"bill" wrote in message t... Ron Capik wrote: bill wrote: Arny Krueger wrote: ....snip.. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. In the limit (per Nyquist) you will not be able to determine the amplitude of that 20k signal with said 40k sample rate. You need to be a little below the Nyquist limit to gain that information, how far below determines how fast you can acquire the amplitude knowledge. Put another way, from the view point of bandwidth limiting, all modulation products need to be below the Nyquist limit else (any out of band) information will be lost. Later... Ron Capik -- The OBVIOUS problem is that if you only have exactly 2 samples per 20KHz sine wave you might sample on the zero crossings and not on the peaks, which would give you an output of 0.0 and not the true amplitude. Simple common sense says that you need to bag at least 4 samples to give a good chance of at least some accuracy. Math can be if not wrong, not optimal in the real world. I live in the real world. Simple common sense overstates the problem by a factor of slightly more than 2. IOW, you can get everything you need to know about phase and amplitude by slightly more than 2 samples per wave. However, if the wave is modulated and its fundamental is too close to the corner frequency of the brickwall filter, then you may lose some information about the modulation. The argument loses a lot of momentum when you consider that most people can't hear over about 15 KHz anyway, and only vaguely over 10 KHz, especially those over 30 years old. Most people I see working at even moderately noisy jobs take no hearing precautions or protection and don't even realize what they have done to themselves until it is already done. Friends of mine who were in bands are now nearly deaf at just over 50, yet I'm 58 and can still hear like I was 15 thanks to avoiding loud night clubs. It is nice to hear the nuances of the higher frequency instruments when some younger people wonder what I am listening to. grin -- Bill (Sleepless biker) Baka |
#340
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
I'm late getting in ths thread, though I see it's still going...
On Mon, 9 Oct 2006 13:25:56 -0400, "Ethan Winer" ethanw at ethanwiner dot com wrote: Ron, I'd be hard pressed to call those lumpy bumpy plots in your paper clear examples of comb filtering. Look again, and you'll see that the blue nulls are more or less evenly spaced, as are the red nulls. Some nulls are much deeper than others, as some "ride on top of" larger changes in level. But the regularly repeating pattern of nulls is pretty clear, with about 11 of them between each vertical major division. It seems to be that "a comb filter" is made by summing two copies of a signal, one delayed in relation to the other. This results in regularly-spaced (frequency-wise) peaks and troughs ("teeth") in the frequency response. In an "average" room, there are (many) more than two copies (reflections) of the signal, each one interfering with each of the others, causing different peak and null frequencies for each of these "combs." I would say the response of the room (or of a reverb unit) is the result of MANY comb filters. So, one can say the response is the result of "comb filtering" but certainly NOT JUST the result of "a comb filter." --Ethan |
#341
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
bill "just use comcast", "you know the drill for no spam bots." wrote:
The OBVIOUS problem is that if you only have exactly 2 samples per 20KHz sine wave you might sample on the zero crossings and not on the peaks, which would give you an output of 0.0 and not the true amplitude. Simple common sense says that you need to bag at least 4 samples to give a good chance of at least some accuracy. Math can be if not wrong, not optimal in the real world. I live in the real world. This stuff was all solved in the mid-1960s, and there is a discussion of just the thing you mention in the FAQ. It's worth reading. Really, the sampling theorem works. I can point you to the math if you really want to see it. And it works in the real world too, as a scope and signal generator will show you. --scott -- "C'est un Nagra. C'est suisse, et tres, tres precis." |
#342
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Arny Krueger wrote:
"bill" wrote in message t... Arny Krueger wrote: "bill" wrote in message m... Arny Krueger wrote: "bill" wrote in message om... Arny Krueger wrote: "bill" wrote in message . com... Arny Krueger wrote: "bill" wrote in message .. . In the days of analog tape, 80 dB dynamic range (without compression) would be truely outstanding performance for analog tape. Today, even consumer audio interfaces have 90 dB dynamic range, and deliver that 90 dB dynamic range over the fully range of audio frequencies with low distortion, which analog tape can't possibly do. Tape had a certain noise floor determined by hiss but the new digital uses sampling techniques with ADC's that are not necessarily linear like the instrument ADC's I used in my work. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. Not true, if the obvious bandwidth restriction and its consequences in the frequency domain are fully considered. More below. As I just mentioned in another post on this what happens if you sample your 20KHz sine wave with a 40KHz sample rate that just happens to sample the zero crossings and not the peaks??? I think you know. I know and also know that any lower frequency than 20 KHz can be fully sampled for amplitude and phase. Your example is one delta (where delta is the smallest non-zero number imagninable) from being false. Not really. The reason that digital gets away from getting called on this is that music has to have more than one cycle to really be heard and almost any instrument puts out a decaying resonant sine at some point. 16 bits is about the limit on true accuracy. No, 19 or 20 bits is pretty common in higher end audio interfaces. But, we're still talking 4 channels in and out for under $1K. The DC accuracy stinks, which I know is not a concern, but check out the INL and DNL figures. Not much of a problem because audio DACs are pretty much all built on the Sigma-Delta model, which inherently has no missing codes and has an inherently monotonic transfer function. It's not practical or reasonable to use the Sigma-Delta model for many instrumention, video and RF applications. Which has been a bummer for me since absolute accuracy is what I need for my instrumentation design work. Low frequency, under 500Hz but it has to be damned accurate. AFAIK Sigma-Delta has no problems with accuracy at low frequencies. It just has implementation problems at very high frequencies. That was a big deal about ten years back but is getting to be less. The Sigma-Delta was a temporary big deal for being able to not see 60 Hz noise in instruments, but I still prefer a SAR ADC. -- Bill (Sleepless biker) Baka |
#343
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Scott Dorsey wrote:
bill "just use comcast", "you know the drill for no spam bots." wrote: The OBVIOUS problem is that if you only have exactly 2 samples per 20KHz sine wave you might sample on the zero crossings and not on the peaks, which would give you an output of 0.0 and not the true amplitude. Simple common sense says that you need to bag at least 4 samples to give a good chance of at least some accuracy. Math can be if not wrong, not optimal in the real world. I live in the real world. This stuff was all solved in the mid-1960s, and there is a discussion of just the thing you mention in the FAQ. It's worth reading. Really, the sampling theorem works. I can point you to the math if you really want to see it. And it works in the real world too, as a scope and signal generator will show you. --scott I do know the math but it can be a problem if you do happen to hit exactly on the zero crossings, math or not. The math is taking into account the ideal of hitting the sample points on the peaks. This is not a problem for music that is so random and not really critical at that frequency. The 44Khz normal sample rate is more than enough for 5 Khz and below where the ear can really make out the difference. Math doesn't lie, but human error can screw up the results. -- Bill (Sleepless biker) Baka |
#344
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
bill "just use comcast", "you know the drill for no spam bots." wrote:
I do know the math but it can be a problem if you do happen to hit exactly on the zero crossings, math or not. That's why you design anti-aliasing filters so that you DON'T hit exactly on the zero crossings. The math is taking into account the ideal of hitting the sample points on the peaks. This is not a problem for music that is so random and not really critical at that frequency. It's not a problem since the anti-aliasing filter prevents anything from being at that frequency. The 44Khz normal sample rate is more than enough for 5 Khz and below where the ear can really make out the difference. Math doesn't lie, but human error can screw up the results. It turns out not to be, since making a sharp enough anti-aliasing filter is difficult. After a decade or so of trying to make sharper and sharper filters with less and less group delay inside the passband, the IC guys started building oversampling filters on a chip. This allows you to sample at a very high sample rate and then low-pass and decimate in the digital domain. Using oversampling means your anti-aliasing filters are easy to build and have a corner frequency way above the frequency of interest. This solves all of the group delay issues and all of the aliasing issues with converters. It doesn't solve the monotonicity issues, but sigma-delta stuff helps with that too. There is a really good description of how this stuff works in the FAQ. --scott -- "C'est un Nagra. C'est suisse, et tres, tres precis." |
#345
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
bill wrote:
Arny Krueger wrote: "bill" wrote in message m... Arny Krueger wrote: "bill" wrote in message om... Arny Krueger wrote: "bill" wrote in message . com... Arny Krueger wrote: "bill" wrote in message .. . In the days of analog tape, 80 dB dynamic range (without compression) would be truely outstanding performance for analog tape. Today, even consumer audio interfaces have 90 dB dynamic range, and deliver that 90 dB dynamic range over the fully range of audio frequencies with low distortion, which analog tape can't possibly do. Tape had a certain noise floor determined by hiss but the new digital uses sampling techniques with ADC's that are not necessarily linear like the instrument ADC's I used in my work. In fact every audio interface I've ever tested (now over 100) has had a nominally flat noise floor. One finds non-flat noise floors in sample format and rate converters. The most reliable difference between instrumentation ADCs and audio ADCs relates to how far they are down at Nyquist, which is a different issue. Instrumentation ADCs tend to be more scupulous about having their response way down at Nyquist. Nyquist is just fine for reconstructing a 20KHz sine wave with a 40KHz sample rate since the filter will output a sine wave, but you really need more samples than that to accurately reconstruct a random audio waveform. Not true, if the obvious bandwidth restriction and its consequences in the frequency domain are fully considered. More below. As I just mentioned in another post on this what happens if you sample your 20KHz sine wave with a 40KHz sample rate that just happens to sample the zero crossings and not the peaks??? I think you know. If you're sampling a 20KHz wave at 40KHz you aren't meeting the criteria of the Nyquist-Shannon Sampling Theorem. The sampling rate has to be GREATER than 2x the frequency of interest. -- Aaron |
#346
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
|
#347
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Scott Dorsey wrote:
bill "just use comcast", "you know the drill for no spam bots." wrote: I do know the math but it can be a problem if you do happen to hit exactly on the zero crossings, math or not. That's why you design anti-aliasing filters so that you DON'T hit exactly on the zero crossings. It still takes a number of sine waves to build up to the desired point, but the filter takes care of rounding the samples into a sine wave. The math is taking into account the ideal of hitting the sample points on the peaks. This is not a problem for music that is so random and not really critical at that frequency. It's not a problem since the anti-aliasing filter prevents anything from being at that frequency. The 44Khz normal sample rate is more than enough for 5 Khz and below where the ear can really make out the difference. Math doesn't lie, but human error can screw up the results. It turns out not to be, since making a sharp enough anti-aliasing filter is difficult. After a decade or so of trying to make sharper and sharper filters with less and less group delay inside the passband, the IC guys started building oversampling filters on a chip. This allows you to sample at a very high sample rate and then low-pass and decimate in the digital domain. This became painfully obvious to me when trying to make a near brick wall filter for some motion control feedback in 2000. Even with 4-8 poles there was no way for a Bessel, Butterworth, Chebyshev, or even Elliptical to do what I needed. I wrote a 'C' program that came close but it sucked up too much of the computer time and I was only trying to have a brick wall at 1.0 KHz. I wound up using some 8 pin digital filter chips that worked just fine and were tunable by changing the filter clock frequency for the switched cap array. Full DSP was not in the cards on this project. Using oversampling means your anti-aliasing filters are easy to build and have a corner frequency way above the frequency of interest. This solves all of the group delay issues and all of the aliasing issues with converters. It doesn't solve the monotonicity issues, but sigma-delta stuff helps with that too. There is a really good description of how this stuff works in the FAQ. --scott Been there, done that, as mentioned above. Good software can be a great thing. -- Bill (Sleepless biker) Baka |
#348
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"bill" wrote in message
t... In the limit (per Nyquist) you will not be able to determine the amplitude of that 20k signal with said 40k sample rate. You need to be a little below the Nyquist limit to gain that information, how far below determines how fast you can acquire the amplitude knowledge. Put another way, from the view point of bandwidth limiting, all modulation products need to be below the Nyquist limit else (any out of band) information will be lost. The OBVIOUS problem is that if you only have exactly 2 samples per 20KHz sine wave you might sample on the zero crossings and not on the peaks, which would give you an output of 0.0 and not the true amplitude. Simple common sense says that you need to bag at least 4 samples to give a good chance of at least some accuracy. Math can be if not wrong, not optimal in the real world. I live in the real world. Simple common sense would tell you that, but sometimes simple common sense is wrong. This is one of those times. Yes, you're correct: 2 samples of a wave won't give you an accurate rendition. That's why you have a steep filter right below the Nyquist frequency (1/2 the sampling frequency). Arguments about whether a 44.1kHz sampling rate is sufficient aren't about whether that rate can accurately sample material up to 20kHz. It can, in theory, provided everything *above* 20kHz is filtered out. (And real-world practice is catching up with theory; the hardware is improving.) The argument, instead, is whether you need a wider bandwidth for your ears to perceive it as sounding right. Not yet resolved. Peace, Paul |
#349
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"bill" wrote in message
t... As I just mentioned in another post on this what happens if you sample your 20KHz sine wave with a 40KHz sample rate that just happens to sample the zero crossings and not the peaks??? I think you know. Which is why you don't do that. Period. Congratulations; you've just discovered aliasing. Peace, Paul |
#350
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
On Wed, 25 Oct 2006 17:59:52 GMT, bill
wrote: As I just mentioned in another post on this what happens if you sample your 20KHz sine wave with a 40KHz sample rate that just happens to sample the zero crossings and not the peaks??? I think you know. Yeah. We "know". But, in this case, I'm told (by people who know their stuff) that we're wrong. NYquist isn't very intuitive. But it works. |
#351
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
If "the scientific method" were the definition of science, then the best
scientists would be those who where the best at running scientific experiments. But despite the need to carefully desingn, run, and interpret experiments, the best scientists are those with the best insight, the best imaginations. The first hypothesis has to come from -- and it doesn't come from experiment. It comes from imagination. |
#352
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
William Sommerwerck wrote:
If "the scientific method" were the definition of science, then the best scientists would be those who where the best at running scientific experiments. But despite the need to carefully desingn, run, and interpret experiments, the best scientists are those with the best insight, the best imaginations. The first hypothesis has to come from -- and it doesn't come from experiment. It comes from imagination. I don't quite know how to evaluate your use of the term "best." Could you quantify that a bit so we're all on the same page? Later... Ron Capik -- |
#353
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
On Thu, 26 Oct 2006 02:01:30 GMT, Ron Capik
wrote: William Sommerwerck wrote: the best scientists are those with the best insight, the best imaginations. I don't quite know how to evaluate your use of the term "best." Could you quantify that a bit so we're all on the same page? A) Seven 2) The best questions. Much thanks, as always, Chris Hornbeck |
#354
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
William Sommerwerck wrote:
If "the scientific method" were the definition of science, then the best scientists would be those who where the best at running scientific experiments. But despite the need to carefully desingn, run, and interpret experiments, the best scientists are those with the best insight, the best imaginations. The first hypothesis has to come from -- and it doesn't come from experiment. It comes from imagination. The theorist and the experimentalist are both essential to the scientific method. Sometimes they are embodied in one person and sometimes it takes many of each to do science. Why can't you see that the imagination and intuition you speak of are integral to the method which defines science both at the level of theory and at the level of experiment? The experiment without a hypothesis is blind and the hypothesis without the experiment is dumb. Neither are science without the other. You can call one Fred and the other Nancy if you want but you can't call them science until they are married. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#355
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
|
#356
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
|
#357
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"Anonymous" wrote in message er.mixmin.net... In article "Arny Krueger" wrote: "Cyberiade.it Anonymous Remailer" wrote in message eriade.it... snip......snip Arny is not a "poor thinker," he's a non-thinker. That really hurts coming from someone who is afraid of his own name. ;-) Nice try, but Bob Cain holds the record to date at having the most lies in a single sentence. That record is three. Better luck next time. Furthermore I have no doubt that Arny lacks the academic credentials necessary to enable him to be rightfully referred to be as a "scientist." If he does, he should contact the graduate school that he attended and ask for a refund. Tell us about your credentials. The issue is whether you, not I, have credentials as a scientist.....and we know the answer about you. What you either know or don't know about about my credentials is irrelvant. Mine are on public record in the Usenet archive. Perhaps you should get it removed until such time as you have accomlished something that is substantive, significant and truly noteworthy. FWIW, the true scientists in my family came up in the next generation. If so, they obviously did not either follow in your footsteps or have you as their mentor. It is amusing how many posters here have to attack the person all the time. It is surely a sign of something amiss when there has to be personal attacks rather the discussion of the relevant points. I think we all know very well qualified people who don't demonstrate the level of common sense and intelligence the qualifications would suggest, and it appears to me that that hypothesis is reinforced by the posters regularly resorting to personal attack. |
#358
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Chevdo wrote:
In article , says... The experiment without a hypothesis is blind That is the WRONG term. No, just a different use of it than yours. Your mind is insufficiently subtle to see the difference. Hasn't school started yet? Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#359
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
"APR" wrote in message It is amusing how many posters here have to attack the person all the time. No one who is *truthfully* regular attacks anyone unless it's political in nature. The rest of it is people just passing through who think they know it all anyway. |
#360
Posted to rec.audio.pro
|
|||
|
|||
I found the holy grail that explains audiophile beliefs
Paul Stamler wrote:
"bill" wrote in message t... In the limit (per Nyquist) you will not be able to determine the amplitude of that 20k signal with said 40k sample rate. You need to be a little below the Nyquist limit to gain that information, how far below determines how fast you can acquire the amplitude knowledge. Put another way, from the view point of bandwidth limiting, all modulation products need to be below the Nyquist limit else (any out of band) information will be lost. The OBVIOUS problem is that if you only have exactly 2 samples per 20KHz sine wave you might sample on the zero crossings and not on the peaks, which would give you an output of 0.0 and not the true amplitude. Simple common sense says that you need to bag at least 4 samples to give a good chance of at least some accuracy. Math can be if not wrong, not optimal in the real world. I live in the real world. Simple common sense would tell you that, but sometimes simple common sense is wrong. This is one of those times. Yes, you're correct: 2 samples of a wave won't give you an accurate rendition. That's why you have a steep filter right below the Nyquist frequency (1/2 the sampling frequency). Arguments about whether a 44.1kHz sampling rate is sufficient aren't about whether that rate can accurately sample material up to 20kHz. It can, in theory, provided everything *above* 20kHz is filtered out. (And real-world practice is catching up with theory; the hardware is improving.) The argument, instead, is whether you need a wider bandwidth for your ears to perceive it as sounding right. Not yet resolved. Peace, Paul It can't, theory or not capture a single sine wave with any accuracy. Sorry, it just won't work. If you can't wrap your head around this then plot it out on a piece of paper and just look. Too many people are brainwashed these days not to think for themselves. -- Bill (Sleepless biker) Baka |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
What are they Teaching | Audio Opinions | |||
Linux is dead...It doesn't even have a pulse. | Pro Audio | |||
HAHA I FIGURED OUT THE holy grail of distorted guitar micing | Pro Audio | |||
Best and Worst in search of the holy grail? | Audio Opinions |