Testing methodology
Hi to all. I have not heard the Bybee treatments, but i would sure like to see how they perform when subjected to tests such as Nordost did on their cables and supply combo.
In short, they showed a new way of measuring which came up with significant differences and pretty much prove that YES they do perform better than zip cord. Please investigate this issue, it could be of similar application with the Bybee stuff.
This info is found at:
1) nordost.com/downloads.asp
2) find New Approaches To Audio Measurement.pdf
Nordost presents an ACTUAL TESTING PROCEDURE using new forms of measurement, for some of the factors that have previously been held to be not measurable or audible.... and which to my eyes look like solid evidence that indeed, some cables, connectors, power conditioners etc. actually DO present significant differences which closely track perceived improvement.
It would be great to see how Bybee performs when tested as shown in the Nordost paper.
My apologies if any one has already brought this up but I haven't been able to read through the whole thread. If this is new information, I hope it helps in your studies.
All the Best.
Hi to all. I have not heard the Bybee treatments, but i would sure like to see how they perform when subjected to tests such as Nordost did on their cables and supply combo.
In short, they showed a new way of measuring which came up with significant differences and pretty much prove that YES they do perform better than zip cord. Please investigate this issue, it could be of similar application with the Bybee stuff.
This info is found at:
1) nordost.com/downloads.asp
2) find New Approaches To Audio Measurement.pdf
Nordost presents an ACTUAL TESTING PROCEDURE using new forms of measurement, for some of the factors that have previously been held to be not measurable or audible.... and which to my eyes look like solid evidence that indeed, some cables, connectors, power conditioners etc. actually DO present significant differences which closely track perceived improvement.
It would be great to see how Bybee performs when tested as shown in the Nordost paper.
My apologies if any one has already brought this up but I haven't been able to read through the whole thread. If this is new information, I hope it helps in your studies.
All the Best.
He's not making it for the military that often mandates the use of bailing wire for resistor leads.
He's making it for audiophiles who don't.
Furthermore, selling them claiming the leads to be copper is fraud.
se
After reading those articles, I'm pretty sure it's a marketing ploy. No wonder people get confused reading this stuff.
Audio tests have included "chirps" and other extremely dynamic (and repeatable) signals. In addition, they are pretty good at controlling all the various situations that can affect the outcome of a test.
Bottom line from the two articles (I read both)? Spend more on cables, like up to 50% of your total budget! Also, clean AC power is paramount. Seems to me that the job of your power supply (normally included when you buy the equipment
) is to isolate the circuits from the garbage on the line.
People, if your equipment has a properly designed power supply, nothing done to the incoming AC power will have an audible effect, short of disconnecting the power that is. Having said that, most devices I see do not have a properly designed power supply.
-Chris
Audio tests have included "chirps" and other extremely dynamic (and repeatable) signals. In addition, they are pretty good at controlling all the various situations that can affect the outcome of a test.
Bottom line from the two articles (I read both)? Spend more on cables, like up to 50% of your total budget! Also, clean AC power is paramount. Seems to me that the job of your power supply (normally included when you buy the equipment
) is to isolate the circuits from the garbage on the line.People, if your equipment has a properly designed power supply, nothing done to the incoming AC power will have an audible effect, short of disconnecting the power that is. Having said that, most devices I see do not have a properly designed power supply.
-Chris
confidence in measurements
No confidence in Nordost? So?
I'm not talking about their marketing hype, I'm talking about the MEASUREMENTS and the METHODOLOGY referred to in my previous post.
the measurements referred to in the Nordost presentation appear to have been done by people who indeed DO seem to know what they're doing.
They also appear quite valid to me. (Yes. I know nothing, but I also have a BSEE. Still ignorant though. Only difference is with enough education it became clearer to me just how much I don't know, which is a lot.)
HEY!
The fact that the Nordost marketing people don't seem to know how to word the info properly in some of their publications should not detract from the validity of the test method, nor of the tests themselves.
It could be useful to not throw out the good with the bad before we've had a good look at what people are doing or judging them for stupid marketing prose.
Otherwise we might be missing the central point..(Hello, Sy?)... which is:
LOOK AT THE METHODOLOGY described on that paper, it might be useful for this specific testing of the Bybee and/or other forms of what suspect to be magic dust.
After all, until they thoroughly read the Nordost paper, and ponder its implications, many of my fellow engineers will probably still "know for sure" that there is no difference in cables.... BUT...
Now that someone has devised a meaningful test procedure that shows significant differences, repeatedly and reliably, under controlled circumstances months apart, well, I wonder if suddenly these same engineers will suddenly be able to hear the difference? (Power of suggestion works in reverse too!)
As for Nordost, I have heard it and it sounded pretty freakin' good when compared to several other cables. Can't buy em yet, too expensive for me.
And I wonder just how much of a change the Bybees make, but will probably never know because it looks like too much expenditure.
That we may not be able to measure a difference with our EXISTING tests, doesn't mean we won't be able to with other methodologies.
And just for the record, my guess is that the Bybees are snake oil, but I am not willing to state that as a truth....'cause I don't know if it is!... and... because in a year or so, there might be some test that shows they are indeed helpful. And I don't like cleaning rotten egg off my face, much less getting it on there in the first place.
No confidence in Nordost? So?
I'm not talking about their marketing hype, I'm talking about the MEASUREMENTS and the METHODOLOGY referred to in my previous post.
the measurements referred to in the Nordost presentation appear to have been done by people who indeed DO seem to know what they're doing.
They also appear quite valid to me. (Yes. I know nothing, but I also have a BSEE. Still ignorant though. Only difference is with enough education it became clearer to me just how much I don't know, which is a lot.)
HEY!
The fact that the Nordost marketing people don't seem to know how to word the info properly in some of their publications should not detract from the validity of the test method, nor of the tests themselves.
It could be useful to not throw out the good with the bad before we've had a good look at what people are doing or judging them for stupid marketing prose.
Otherwise we might be missing the central point..(Hello, Sy?)... which is:
LOOK AT THE METHODOLOGY described on that paper, it might be useful for this specific testing of the Bybee and/or other forms of what suspect to be magic dust.
After all, until they thoroughly read the Nordost paper, and ponder its implications, many of my fellow engineers will probably still "know for sure" that there is no difference in cables.... BUT...
Now that someone has devised a meaningful test procedure that shows significant differences, repeatedly and reliably, under controlled circumstances months apart, well, I wonder if suddenly these same engineers will suddenly be able to hear the difference? (Power of suggestion works in reverse too!)
As for Nordost, I have heard it and it sounded pretty freakin' good when compared to several other cables. Can't buy em yet, too expensive for me.
And I wonder just how much of a change the Bybees make, but will probably never know because it looks like too much expenditure.
That we may not be able to measure a difference with our EXISTING tests, doesn't mean we won't be able to with other methodologies.
And just for the record, my guess is that the Bybees are snake oil, but I am not willing to state that as a truth....'cause I don't know if it is!... and... because in a year or so, there might be some test that shows they are indeed helpful. And I don't like cleaning rotten egg off my face, much less getting it on there in the first place.
Actually, the data presented in the Nordost deal didn't really indicate that there was a difference beyond test repeatability. Hard to say because the graphs are so low res, you can't read the scales and they presented no data showing run to run repeatability. Some of the graphs looked like there was binning and windowing error- maybe deliberate, maybe incompetence, I have no idea. This was clearly written for marketing purposes and clearly by either someone who didn't know what he was doing or someone who does but assumes his audience doesn't.
Having tried very hard to do the "subtraction" tests shown in the Nordost doc, I can tell you that they are fiendishly difficult. At least they admit as much in the doc. Getting the amplitudes exact and the the time right to the sample is much harder than you might think.
In fact I don't see how it can be done without a synchronized DAC and ADC. Even on short, simple sound files it's no walk in the park.
Too bad, as the idea holds a lot of promise. Just can't tell from that document what they did to overcome the speed bumps. They do seem to have taken them into consideration, but I don't see enough to be confident that things line up that well.
In fact I don't see how it can be done without a synchronized DAC and ADC. Even on short, simple sound files it's no walk in the park.
Too bad, as the idea holds a lot of promise. Just can't tell from that document what they did to overcome the speed bumps. They do seem to have taken them into consideration, but I don't see enough to be confident that things line up that well.
Hi Jack,
In addition to what SY said, there is a lack of detail on exactly what equipment was used. This can make a difference as any short coming in it's design can be affected by an external device. A power supply especially! There wasn't enough evidence presented to come to any conclusions. As I mentioned, with a defect design in equipment, you can make a difference depending on the connections and purity of the applied AC mains.
I can't say whether their cables are good or bad. What I can say is that there are probably situations where cable type and construction can "fix" a problem area.
As to test methodologies, the capture and subtract trace is a common procedure. I do it all the time with both an oscilloscope and spectrum analyzer. Those functions are built right in to the equipment. Nothing new or earth shattering about the tests they were running.
-Chris
In addition to what SY said, there is a lack of detail on exactly what equipment was used. This can make a difference as any short coming in it's design can be affected by an external device. A power supply especially! There wasn't enough evidence presented to come to any conclusions. As I mentioned, with a defect design in equipment, you can make a difference depending on the connections and purity of the applied AC mains.
I can't say whether their cables are good or bad. What I can say is that there are probably situations where cable type and construction can "fix" a problem area.
As to test methodologies, the capture and subtract trace is a common procedure. I do it all the time with both an oscilloscope and spectrum analyzer. Those functions are built right in to the equipment. Nothing new or earth shattering about the tests they were running.
-Chris
Hi panomaniac,
Some post processing software allows you to align the curves. Certainly not a real-time thing though! Also, an FFT type record is pretty easy since the frequencies will (should) be the same. A time record is the more difficult case, and a time record with math done would be horrendous to do!
As I said, I commonly do this with a spectrum analyzer.
-Chris
Some post processing software allows you to align the curves. Certainly not a real-time thing though! Also, an FFT type record is pretty easy since the frequencies will (should) be the same. A time record is the more difficult case, and a time record with math done would be horrendous to do!
As I said, I commonly do this with a spectrum analyzer.
-Chris
Capture and subtract is indeed possible on a scope if the two files are correlated, or captured at the same time. Capture and subtract after the fact on two different wave files that were not captured simultaneously is extremely difficult. If the clocks for the two files were not syncronized, there is no way that the samples will contain the same information. The error may be small, but it will vary every time you run the test.
I have been working on a way to start with a wave file ripped from a CD, play it in a high end DAW (Cakewalk Sonar) through a DUT using a 24 / 192 sound card, and capture the DUT output in the DAW. The two wave files can be correlated in time to compensate for delay and subtracted to show the error. This is still not perfect and the error is non zero and not repeatable with a plain wire under test. The errors are getting small though.
Test methods
Gentlemen:
Outside of the scope of their paper, in some of their latter announcements, Nordost claim that they ran the same tests on the same equipment months later at a different location, and the error differences showed up in exactly the same places. That is significant.
Yes I know, where's the peer review, etc.... but, guys... I've spent some time in the test equipment and signal analysis business, both as a user and apps engineer... and I do understand how to derive a difference signal using single shot storage scopes, repetitive and pulse averaging on HF test gear, using complex waveforms, studying reflections w high speed signal analyzers, etc. and applied it extensively in various fields of endeavor from medical to aerospace to low frequencies, including audio. And we extensively used signal subtraction and comparison with in real time and with repetitive functions. Because I had to deal with so many types of applications I became somewhat familiar with different testing methodologies and approaches.
And it is my feeling that there is something in this test methodology that merits more than a casual dismissal especially IF as Nordost says: the tests run at a different time in a different location matched exactly with the originals... the errors correlated, that is, they happened in the same places and with similar amplitudes....
well, uh, who cares about the scale? If it's small and repeatable and it CORRELATES with what we are HEARING, it means we are finding something we were previously MISSING in our measurements, no?
Please do not dismiss this method, I believe it has some real potential.
Yes Chris, I believe your power supply comments are correct, when it comes to power supplies, very few are any good and it is easy to see differences here.
So, if Acuity's results were based only on the PS, I would say you are totally correct. But the tests were done in sequence, with each element and then as a whole....
and they stated they showed differences for EACH of interconnects, speaker cables, power chord, power conditioner and platform. And according to them each showed a consistent difference.
SO... before we dismiss this type of testing because it's difficult to get proper mathematical correlation (Notice the wording: I said difficult, not impossible, apparently.)
How about checking with the guys that invented the methodology to see if they would volunteer some info on how to do this (not too likely if it's proprietary code), or even get a group funding together to have him check on a Bybee?
Here's the guy:
Dr Gareth Humphries-Jones of Acuity Products, a Doctor of Applied Mathematics and a specialist in signal processing algorithms as applied to high performance sonar and radar systems).
Wanna go out and check on them? I did, here's Gareth's info:
Gareth Humphreys-Jones - United Kingdom | LinkedIn
and here...
Background
If I get the gist of it, he's a real whiz with mathematical processing and signal analysis and he has loads of aerospace contracts.
So IF what they said about their test method actually happens to be true... and it looks very interesting to me... the Bybee and similar boxes would be good candidates to check out and see if they have similar effects to the Nordost gear... and maybe we can all benefit from learning more about this.
Gentlemen:
Outside of the scope of their paper, in some of their latter announcements, Nordost claim that they ran the same tests on the same equipment months later at a different location, and the error differences showed up in exactly the same places. That is significant.
Yes I know, where's the peer review, etc.... but, guys... I've spent some time in the test equipment and signal analysis business, both as a user and apps engineer... and I do understand how to derive a difference signal using single shot storage scopes, repetitive and pulse averaging on HF test gear, using complex waveforms, studying reflections w high speed signal analyzers, etc. and applied it extensively in various fields of endeavor from medical to aerospace to low frequencies, including audio. And we extensively used signal subtraction and comparison with in real time and with repetitive functions. Because I had to deal with so many types of applications I became somewhat familiar with different testing methodologies and approaches.
And it is my feeling that there is something in this test methodology that merits more than a casual dismissal especially IF as Nordost says: the tests run at a different time in a different location matched exactly with the originals... the errors correlated, that is, they happened in the same places and with similar amplitudes....
well, uh, who cares about the scale? If it's small and repeatable and it CORRELATES with what we are HEARING, it means we are finding something we were previously MISSING in our measurements, no?
Please do not dismiss this method, I believe it has some real potential.
Yes Chris, I believe your power supply comments are correct, when it comes to power supplies, very few are any good and it is easy to see differences here.
So, if Acuity's results were based only on the PS, I would say you are totally correct. But the tests were done in sequence, with each element and then as a whole....
and they stated they showed differences for EACH of interconnects, speaker cables, power chord, power conditioner and platform. And according to them each showed a consistent difference.
SO... before we dismiss this type of testing because it's difficult to get proper mathematical correlation (Notice the wording: I said difficult, not impossible, apparently.)
How about checking with the guys that invented the methodology to see if they would volunteer some info on how to do this (not too likely if it's proprietary code), or even get a group funding together to have him check on a Bybee?
Here's the guy:
Dr Gareth Humphries-Jones of Acuity Products, a Doctor of Applied Mathematics and a specialist in signal processing algorithms as applied to high performance sonar and radar systems).
Wanna go out and check on them? I did, here's Gareth's info:
Gareth Humphreys-Jones - United Kingdom | LinkedIn
and here...
Background
If I get the gist of it, he's a real whiz with mathematical processing and signal analysis and he has loads of aerospace contracts.
So IF what they said about their test method actually happens to be true... and it looks very interesting to me... the Bybee and similar boxes would be good candidates to check out and see if they have similar effects to the Nordost gear... and maybe we can all benefit from learning more about this.
No one is dismissing the method out of hand. I think that's unfair. We have simply noted that it's rather difficult to do. And the the Nordost paper does not supply enough detail to determine what steps were taken to overcome the difficulties. It's not so much the testing, as it is the reporting.
Please feel free to contact Dr Humphreis-Jones. There were a number of folks here willing to put $ into the pot for Bybee purchase, perhaps the same would go for testing it we are satisfied the tests are worthwhile.
Please feel free to contact Dr Humphreis-Jones. There were a number of folks here willing to put $ into the pot for Bybee purchase, perhaps the same would go for testing it we are satisfied the tests are worthwhile.
I've released freeware software to do such alignment and subtraction of audio time domain wav files (along with the recording functions, and even sample rate matching, and an option for frequency response equalization). It's been brought up from time to time in these forums, but for those not familiar with it --
"Audio DiffMaker" (freeware)
Audio DiffMaker
It operates in both time domain and frequency domain, as most appropriate for each function. And at sample rates to 192kHz and resolutions to 24bit. Maybe this would be another good way to check the Bybee filters with.
- Status
- Not open for further replies.