 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
In Reply to: Re: John, you were sucker-punched, then piled on, by the Usual Suspects. posted by Steve Eddy on June 09, 2003 at 16:35:25:
I would say: 50 dB in this case.
Follow Ups:
I would say: 50 dB in this case.You would say? You don't seem too sure of that. Why would it apply 50dB of gain? I went back and searched a previous thread and found where you said that the oscillator had a fixed output of about 3 volts. To get the 30mV output you fed into the cable, that would require 40dB of attenuation.
If the 1700B's supposed to be normalizing the gain, why would it apply more gain than attenuation? I thought the point of normalizing the gain was so that the magnitude of the harmonics would be the same as they would be if they were referenced to the 3 volt output of the oscillator? In that case, it should be applying as much gain as attenuation. Which in this case it would be 40dB.
So how sure are you about the 50dB figure? If the 3 volt fixed level of the oscillator is correct, and the 30mV signal you sent through the cable is correct, 50dB doesn't seem to make any sense. Does the notch filter have a 10dB insertion loss or something?
se
Please ! Just add 50 dB to the graphs and you will be within 2dB.
Please ! Just add 50 dB to the graphs and you will be within 2dB.Within 2dB of what? Why the 50dB? What sense does it make? What exactly is being "normalized" by 50dB of gain? Do you know what your 1700B is actually doing or don't you? Are you just guessing? What?
se
Steve, please ask someone else about the Sound Technology 1700B. Perhaps they can explain its operation to you. I have used the unit for the last 27 years. I have some idea of its operational capabilities, but apparently you haven't ever used an analyzer of this type. I am optimistic that you actually want answers to your questionds, but sometimes I wonder.
In any case, the distortion output of the Sound Technology is set for -30dbV for -80dB harmonic distortion. This is factory set. This output also reads -80dB on the distortion scale. Therefore -45dB would be 15db lower than -80dB, which is -95dB. This is the distortion residual of the ST1700B. The harmonics are present at a somewhat lower level on my graphs. They can only be seen and measured with a spectrum analyzer which separates the spectrum into its harmonics. The spectrum analyzer also removes extra noise added by the relatively large bandwidth.
This can be done in two ways: One approach is by using a narrow resolution bandwith of 10-30 Hz or so. This will lower the noise so that the harmonics can be seen.
A second way is to use signal averaging with an FFT analyzer. In this case I use 100 or more measurements and add them together electronically. The harmonics add directly, but the noise adds randomly, so it doesn't increase as fast as the signal. This also separates the distortion residual from the noise.
The absolute magnitude that Steve Eddy is concerned about is really not very important, because I do RELATIVE testing between different cables, and this does not require exact calibration of the distortion harmonics, just a RELATIVE equivalance in the measurement of the distortion harmonics of the cables. I hope that any of you interested in this test, now have some idea of how it is conducted.
The absolute magnitude that Steve Eddy is concerned about is really not very important, because I do RELATIVE testing between different cables, and this does not require exact calibration of the distortion harmonics, just a RELATIVE equivalance in the measurement of the distortion harmonics of the cables. I hope that any of you interested in this test, now have some idea of how it is conducted.It's important because I want to know EXACTLY where you're measuring. I don't want to guess and I don't want your guesses. The reason I want to know EXACTLY where you're measuring is because I don't want any subsequent measurements which may not turn up the same results as yours be dismissed by your saying that they weren't measuring far enough down. If we know EXACTLY where you're measuring, and know EXACTLY where any subsequent measurements are measuring, then there can be no arguing about the results.
In any case, if we take your word about adding 50dB to your plots to get the actual levels being measured, that means that your noise floor is bottoming out at -120dBV (70 + 50).
If we take a look at Arny's measurements, we see that his noise floor is bottoming out at -138dBV. Which means that he's measuring another 18dB below your measurements.
And if you take his baseline measurement of a LynxTWO XLR cable looped through from output to input and overlay the measurement of the LynxTWO plus two 12 foot lengths of Radio Shack cable strung together for a total of 24 feet, you see virtually no difference at all in the magnitude of the distortion products.
The only increases are a slight increase in the noise floor at lower frequencies and an increase in the spurious signals such as power line harmonics and a few other non-harmonically related spikes.
And keep in mind that these measurements were made on a PC-based system made by the same company that made your 1700B.
se
Steve, and anyone else interested, please look at Arnies graphs closely. This is a ONE stage measurement, and a very good one in general. I use a TWO stage measurement, and this gives me a technical advantage.
OK folks, what is the magnitude of the 7th harmonic on Arnies second graph? Do you think that it is -125dB ? Well, that's wrong, as you have to subtract 30 db from it, because the graph does NOT start from the top That gives -95dB for the seventh harmonic. My RS case 7th harmonic in my test is about -110 to -115db dB. My residual with a VDH cable is usually about 120 to -125dB.
Remember, the important thing to consider here is the 30 dB offset in Arnies measurements. Distortion products must be measured against the reference signal, in Arnies case, but my ST1700 does the normalization automatically, so I don't have to change my measurements to get the correct result.
Steve, and anyone else interested, please look at Arnies graphs closely. This is a ONE stage measurement, and a very good one in general. I use a TWO stage measurement, and this gives me a technical advantage.
OK folks, what is the magnitude of the 7th harmonic on Arnies second graph? Do you think that it is -125dB ? Well, that's wrong, as you have to subtract 30 db from it, because the graph does NOT start from the top That gives -95dB for the seventh harmonic. My RS case 7th harmonic in my test is about -110 to -115db dB. My residual with a VDH cable is usually about 120 to -125dB.
Remember, the important thing to consider here is the 30 dB offset in Arnies measurements. Distortion products must be measured against the reference signal, in Arnies case, but my ST1700 does the normalization automatically, so I don't have to change my measurements to get the correct result.
Steve, and anyone else interested, please look at Arnies graphs closely. This is a ONE stage measurement, and a very good one in general. I use a TWO stage measurement, and this gives me a technical advantage.What advantage, John? The only difference is that you're nulling out the fundamental and applying some voltage gain. What advantage does that give you? You're not changing the ABSOLUTE values. You're just playing decbiel games.
OK folks, what is the magnitude of the 7th harmonic on Arnies second graph? Do you think that it is -125dB ? Well, that's wrong, as you have to subtract 30 db from it, because the graph does NOT start from the top
No, you don't have to substract 30dB from it. That's the nice thing about Arny's plots. You don't have to do ANYTHING to know what's going on, unlike yours where you don't seem to have the foggiest idea how just much gain is being applied AFTER the cable and BEFORE your Mac The Scope so no one can look at your plots and be able to determine EXACTLY what the ABSOLUTE values are.
Your Mac The Scope's dB scale is referenced to 1 volt, same as Arny's SpectraLAB. Arny fed the signal straight into the cable and straight from the cable back to the analyzer. You run yours through a notch filter and God knows how much voltage gain so that no one can determine ABSOLUTE values for your plots.
If your claim that one needs to add 50dB to your plots to get the proper values is true, then YOUR 7th harmonic is indeed at -113dBV and Arny's is indeed at -125dBV.
That gives -95dB for the seventh harmonic. My RS case 7th harmonic in my test is about -110 to -115db dB.
No, John. That gives -125dBV for Arny's seventh and -112dBV for your seventh.
My residual with a VDH cable is usually about 120 to -125dB.
John, if your 50dB figure is to be believed, your noise floor BOTTOMS OUT at -120dBV. So is the 50dB figure you claimed incorrect or what? Do you have the foggiest idea what you're doing?
Remember, the important thing to consider here is the 30 dB offset in Arnies measurements.
No, John. There isn't a 30dB offset in Arny's measurements. His measurements were taken straight off the end of the cable as it was being fed a 30mV signal.
What there is is a 50dB offset in YOUR measurements. That's because you're APPLYING VOLTAGE GAIN BETWEEN THE OUTPUT OF THE CABLE AND THE INPUT IF YOUR MAC THE SCOPE! But you don't seem to REALLY know EXACTLY how much gain is being applied so no one can look at your plots and know EXACTLY what the ABSOLUTE levels are.
Distortion products must be measured against the reference signal, in Arnies case, but my ST1700 does the normalization automatically, so I don't have to change my measurements to get the correct result.
And in Arny's case, the reference signal is left intact so we can all see EXACTLY what's going on. Because you're nulling out the fundamental and applying some amount of voltage gain, your plots are largely worthless because no one can know what the ABSOLUTE levels are without relying on your knowing how much voltage gain is being applied.
Here's the facts, John:
If what you say is true and one has to add 50dB to your plots to get the proper dBV levels, then Arny's noise floor is at -138dBV and yours is at -120dBV. Making Arny's noise floor 18dB lower than yours. And Arny's 7th harmonic is at -125dBV, while yours is at -112dBV. Making Arny's 7th harmonic 13dB lower than yours.
Are you just playing a smoke and mirrors decibel game here or do you simply not understand what's going on?
se
Steve, I don't know who you are or who you work for, but I find your attitude offensive.There are a lot of people that are so insecure they don't trust their own senses (perhaps that makes them "senseless") and claim that all cables *must* sound the same because they can't measure any difference between them. John Curl has come up with an interesting measurement technique that shows differences between cables. This is an extremely important finding. Furthermore it is *valid* because it is *repeatable*. You could find this out for yourself if you cared to. You can get a used Sound Technology on e-Bay for a couple of hundred dollars, and I would assume that the sound cards are about the same price.
Instead, you seem to be hung up on the "gain" that the Sound Tech applies to its residual (nulled) output. As if this meant anything important! John has already told you what the signal level he applied to the DUT. If you were interested in learning something, you could pick up the ball and run with it. But it seems you are only intent upon building yourself up by tearing others down...
You have clearly ("EXACTLY"?) demonstrated your lack of understanding by missing areas of even greater uncertainty, such as the averaging technique used. Nothing could be quantified without knowing the exact number of measurements that were being averaged. But that is not the point of this test. The point is that John has found a way to measure differences between interconnect cables. Furthermore, that measurement is repeatable and is *not* due to artifacts created by the test setup.
You seem incapable of grasping the importance of this. If you want to refine the results (certainly a worthwhile goal), please do so. But don't whine that John doesn't do his work the way that you think he should do it. Please feel free to extend his work. (As I said, a worthwile goal.)
The thing I don't understand is why people with this attitude came here. I used to hang out at the rec.audio newsgroups until this poor attitude made the signal-to-noise ratio too low to be useful. When I came to the Asylum things were noticeably better. Now we seem to be in a decline back to the same type of problems that plagued rec.audio.
Charles Hansen
Thanks Charles for coming to bat for me. The criticisms on this topic by SE and several others just shows a lack on their part to understand what we can measure, and their inability to actually measure anything with any sophistication. It's kind of sad, as I expected better from them. Rest assured, I have made hundreds of similar measurements.
Steve, I don't know who you are or who you work for, but I find your attitude offensive.I'm me. And I work for myself as well as a company that you once approached to design and build amplifiers for you.
There are a lot of people that are so insecure they don't trust their own senses (perhaps that makes them "senseless") and claim that all cables *must* sound the same because they can't measure any difference between them.
I trust my senses completely with respect to telling me what I like or don't like. I don't trust my senses completely with respect to establishing physical objective realities. This has nothing to do with insecurity but rather some 100 years of research which has shown time and again that our subjective perceptions are not always unerring reflections of physical realities. The ones who are insecure are the ones who are in complete denial of this fact and can't handle the notion that they're just mortal human beings subject to the same flaws and weaknesses as other mortal human beings.
As for claiming that all cables *must* sound the same, that is a claim I have NEVER made my life. And if you want to get on my shitlist, the best way to do it is to put words in my mouth and claim I have said things I have never said and hold beliefs that I do not hold. That's what your friend Curl has been doing here for several years and why I have such little respect for him. So shall I assume, based on your comment above, that you will be following suit and claim I have said things I have never said and hold beliefs that I do not hold?
John Curl has come up with an interesting measurement technique that shows differences between cables.
What interesting measurement technique? There's absolutely nothing new or unique about the technique he's using.
This is an extremely important finding.
We don't yet know exactly WHAT he's found. Only what he's claimed to have found. Just because someone has claimed to have found something doesn't necessarily make it so. Need I remind you of Pons and Fleischman?
Furthermore it is *valid* because it is *repeatable*.
Repeatable by whom? John? That doesn't make it valid. Who ELSE beside John has repeated it with the same result? Again, need I remind you of Pons and Fleischman?
You could find this out for yourself if you cared to. You can get a used Sound Technology on e-Bay for a couple of hundred dollars, and I would assume that the sound cards are about the same price.
Why would I want to do that? If the different levels of distortion John is measuring turns out to be nothing more than varying levels of distortion in his 1700B, why would I want to use the same device? If we're to rule out any anomalies due to the 1700B, then it would only make sense to try and repeat the measurements on something other than the 1700B.
Instead, you seem to be hung up on the "gain" that the Sound Tech applies to its residual (nulled) output. As if this meant anything important! John has already told you what the signal level he applied to the DUT.
It's important in order to determine ABSOLUTE levels. Knowing what the signal level applied to the DUT was does NOT establish any absolute levels once the output has been sent through a notch filter and some amount of voltage gain.
John said to add 50dB to his plots to get the proper levels which if that's the case, his measurements aren't measuring as far down as Arny Kruger's measurements and if Arny was measuring what he said he was measuring (two 12 foot Radio Shack interconnects strung togheter) then his measurements show no increase in distortion when 24 feet of Radio Shack interconnects are added. Which means that up to this point, no one else has been able to verify John's results.
You have clearly ("EXACTLY"?) demonstrated your lack of understanding by missing areas of even greater uncertainty, such as the averaging technique used. Nothing could be quantified without knowing the exact number of measurements that were being averaged. But that is not the point of this test. The point is that John has found a way to measure differences between interconnect cables. Furthermore, that measurement is repeatable and is *not* due to artifacts created by the test setup.
Again, who else has repeated John's measurements? And how has it been established with certainty that the measurements are not due to the 1700B? If they are due to the 1700B, John can sit there all day repeating the same results. So what? I'm sure Pons and Fleischman could sit there all day and repeat their results as well.
What counts is INDEPENDENT verification or falsification.
You seem incapable of grasping the importance of this.
And you seem incapable of grasping the importance of INDEPENDENT verification or falsification.
If you want to refine the results (certainly a worthwhile goal), please do so. But don't whine that John doesn't do his work the way that you think he should do it.
Did you actually read any of my posts? Or did John just call you and say "Waaaaaaaaa! Steve's pickin' on me again!"?
If you'd have actually read this exchange, you'd see that I wasn't talking about the way John is doing his work, I was simply trying to found out WHAT his work actually was so I would be able to take his plots and be able to determine the ACTUAL levels they represent. I did this by asking a very simple question. That being exactly how much gain was being applied between the output of the cable and the input of his Mac The Scope.
This was a simple question requiring a simple answer. And if John knew his 1700B as well as he should after having owned it for 27 years, his answer should have been simply "xxdB". But the answers he's given so far indicate that he's only just guessing and doesn't actually know for sure.
He finally managed to come up with a figure of 50dB. And if that's the case, then he's not measuring as far down as he claims to be and that Arny's measurements are measuring significantly further down than John and if Arny was measuring what he said he was measuring, adding 24 feet of Radio Shack interconnects does not add any extra distortion.
The thing I don't understand is why people with this attitude came here.
My only attitude is to get at the truth, wherever it may lay. What I don't understand is the attitude of those who unquestioningly accept anything and everything that comes down the pike.
se
Most of what you posted is a rehash of the same things you've been posting on this topic, which is what I basically expected. But one thing you said genuinely surprised me:"And I work for myself as well as a company that you once approached to design and build amplifiers for you."
I think you must be thinking of someone else. I have never asked another company to "design and build amplifiers" for me.
Charles Hansen
And even though you have not said a thing about cables one way or the other, you are likely somebody who has exotic cabling and has heard differences and improvements over "regular" cables.So if I'm correct in my assessment (and I have not searched for your name in the archives and it is the first time I have seen it), then your post is biased towards John because your beliefs.
I believe you have demonstrated a good example of confirmation bias which is the tendency to put greater weight on evidence that tends to confirm your beliefs and more easily dismiss evidence against.
Don't worry, it happens a lot around here.
You do know about the inquiry function, where you click on the (A) or whatever letter follows the moniker?
My intent was to guess what his beliefs are based on his observations in his posting. A truly unbiased, objective viewpoint does not reveal this. Therefore since I correctly guessed his position, I must have done it by the bias in his post.I see too on his website that he has invented yet another breakthrough in interconnect technology. There seem to be more breakthroughs in audio cable technology these days than you can count dollars in your wallet.
You do know about the inquiry function, where you click on the (A) or whatever letter follows the moniker?Apparently Charles doens't know about it either or he wouldn't have said he doesn't know who I am or who I work for. So don't be too hard on Mike. :)
se
It strikes me as a little too precious and rhetorically over-the-top to feign naivete on this site and this forum.
There are a lot of people ........claim that all cables *must* sound the same because they can't measure any difference between them""CHI do not recall Steve saying that...ever..
""Furthermore it is *valid* because it is *repeatable*.""CH
Reports of repeatable measurements do indeed bolster an argument..However, repeating the same measurement using the same equipment verifies only that the measurement technique produces the same result..That should not be confused with validity..
""You can get a used Sound Technology on e-Bay for a couple of hundred dollars, and I would assume that the sound cards are about the same price.""CH
I believe there's an analyzer in there somewhere...not very cheap, they be...
""Instead, you seem to be hung up on the "gain" that the Sound Tech applies to its residual (nulled) output. As if this meant anything important!""CH
The gain applied to the residual determines what is measured by the analyzer..Assigning numbers such as -120dB down requires absolutely knowing the gain that was applied to bring the residual up to a level that the analyzer is capable of seeing..For Steve to duplicate John's test, that is all important..
""missing areas of even greater uncertainty, such as the averaging technique used""CH
That's further down the road, usually canned into the software or hardware..not even a big thing yet..
""The point is that John has found a way to measure differences between interconnect cables. Furthermore, that measurement is repeatable and is *not* due to artifacts created by the test setup""CH
John has indeed stated such..repeatedly..But has not provided sufficient details to enable anybody else to independently verify..apparently the equipment list by itself is not enough information..And, without equipment details, or independent verification, how is anybody to believe it's not a shield issue, a contact issue, or any other thing that can bugger the measurement of a 30 nanovolt signal..
I have been watching the thread develop, and can understand why Steve asks his questions..He's not received enough to repeat the test..
""You seem incapable of grasping the importance of this.""CHI think that is an incorrect statement w/r to Steve..
""If you want to refine the results (certainly a worthwhile goal), please do so.""CH
It is not possible to refine any results when the basis for the results is unknown..
""But don't whine that John doesn't do his work the way that you think he should do it. Please feel free to extend his work. (As I said, a worthwile goal.)""CH
One cannot extend another's work without knowing how the work was done..Yes, a worthwhile goal, but not possible without details..
John has said "do the work for yourself". But how is that possible without knowing what the details are?
I found one typo in my last input that could be confusing: Please change [-45dB] to -45dbV in this case. dB in this case is a relative measurement of distortion. dbV in this case is a measurement of absolute voltage.
45dbV is the magnitude of the IkHz voltage measured of the nulled fundamental from the distortion output jack. This is the voltage that sets the level on the graphs.
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
