Thanks for that blast from the past, I had completely forgotten 'the blind spot'. Careful where you aim your examples though, what does an experiment that demonstrates perception can block something that really exists say in the current context? Being fair, many visual illusions demonstrate the mind's ability to synthesis things that don't exist, such as the 2-D patterns which pop a third dimension when viewed correctly, or my favourite, the brilliant blue/green circular pattern which appears to rotate on the page.
Johan, perhaps jlsem meant the lack of references in this debate to incontestably solid work in the field.
Johan, perhaps jlsem meant the lack of references in this debate to incontestably solid work in the field.
Johan,
The reason he is challenged is very simple.
Sy stated
That is one hell of an unqualified statment to make, the key words being easily and ever.
Even if we assume that SY’s test methodology is capable of grading 12AT7 in terms of their “goodness” (for want of a better word), when questioned, he won’t reveal what other tubes he tested to arrive at this conclusion, the number of samples he tested or the range of operating conditions over which they were tested. His evasiveness leads me to strongly suspect that the data are insufficient to support the claim.
I’ll bet $10k to James Randi’s left testicle that he wouldn’t be game to submit a paper containing his dataset and conclusions to a refereed journal.
Where I come from, making unqualified statements without the data necessary to support them or the willingness to provide the data for peer scrutiny is plain BAD SCIENCE.
If you wan't an example of how to present experimental findings in a manner that does not invite challenge and ridicule, refer to George's post #450.
I beg to differ. I posted all of this in my original posts about ECC81 tests a couple of years ago. No need to rehash it in a thread about power cords, the search function on the forum will turn it up for you.
Johan, you don't seem to have read my post very carefully. I make no mention of the word truth anywhere in it (nor in any of my other posts in this thread). I didn't mention scientific facts because they are well known. You can measure a cable as accurately as possible and there is no doubting the frequency response of said cable derived from these measurements. These are scientific facts. But if you go on to say that people are being deceived because they hear a difference between two cables despite of these scientific facts, you have formed a scientific hypothesis that needs to be proven before it becomes a truth. And, please, don't remind me that the burden of proof falls on the other side.
Surely you don't believe a logical argument is capable of establishing a scientific fact on its own.
John
Surely you don't believe a logical argument is capable of establishing a scientific fact on its own.
John
I considered it an example of the fact that the mind synthesizes things that aren't there. Even when I cover one eye, I don't notice a blind spot. Wouldn't have ever known without that experiment. I would have assumed that I didn't have one. In fact, I would have insisted. I probably would have gotten mad and called someone ugly names and told them that their JJ 12AT7 tubes sucked if they had told me I had a blind spot.
What it all boils down to is that the senses are measurement systems just as all my lab equipment is. If I am to take valid or meaningful measurements I must understand the limitations of my equipment. What makes the senses most dangerous as test instruments is the fact that they, with the help of the brain, fool us into thinking that they are better than they are. My eyes tricked me into thinking that they were showing me everything in front of me for years.
Hi rdf,
I used to be authorized warranty for that line as well. I remember these. They were a fun little box.
Hi mach1,
Why on earth are you pestering SY??
The man mentioned the equipment he used. Perhaps you aren't aware that an HP 3581A is a tuned frequency wave analyzer. He can tune in and read the amplitude of each harmonic individually. That's better than a THD meter. The easiest way to capture the screen info is to write down the readings unless you have an HP1B bus set up. Also, I have seen this instrument in his lab. It actually exists and works. He even told you were to find the information you requested!
I think you need to prove yourself now since you seem intent on badgering SY here. All you have done to date is challenge him.
Tell me, how do you determine the suitability of components or vacuum tubes? Please state what types made up your test population.
Pretty please?
If you can't provide solid answers, it's high time you drop your incessant picking away at another member.
-Chris
Thanks. I needed that. 🙁
I used to be authorized warranty for that line as well. I remember these. They were a fun little box.
Hi mach1,
Why on earth are you pestering SY??
The man mentioned the equipment he used. Perhaps you aren't aware that an HP 3581A is a tuned frequency wave analyzer. He can tune in and read the amplitude of each harmonic individually. That's better than a THD meter. The easiest way to capture the screen info is to write down the readings unless you have an HP1B bus set up. Also, I have seen this instrument in his lab. It actually exists and works. He even told you were to find the information you requested!
I think you need to prove yourself now since you seem intent on badgering SY here. All you have done to date is challenge him.
Tell me, how do you determine the suitability of components or vacuum tubes? Please state what types made up your test population.
Pretty please?
If you can't provide solid answers, it's high time you drop your incessant picking away at another member.
-Chris
Speaking of cables and scientific facts, I once had a bad time because of a cable in audio.
In 1981 I moved because of family reasons to a town where jobs for electronic engineers in development area were not available because there were no electronics developments in the area, so I had to accept a position in technical support of railway communications and electronics.
However, it was boring, especially to fix things that existed long before I come there, especially when no modern resources I got used to were available.
One of problems I faced, was a problem with phone ringing on one station.
It was a dispatchers' line, 200 miles long pair of wires on poles, to which many high input/output impedance peaces of audio equipment were connected including special ring receivers that reacted on combination of tones in audio band. Everybody who was connected to the line had to hear each other, but they could be called by sending corresponding sets of tones on which thir equipment must react. That particular station reacted on it's tone set intermittently, and support engineers were endlessly balancing from no rings to false alarms...
All modules, including amplifiers with AGC, filters, demodulators, were replaced and readjusted many times, but it did not help.
When I went there I've found that there is a notch exactly on one frequency of the set officially assigned to call this station, and exactly in that location! The line was properly terminated on both ends. Line taster showed reflections, but on the distance of many miles from that troublesome station. However, it was a good clue, and I've found that as the result of reconstruction on the station in the distance of many miles the line was cut and a peace of underground cable was fitted in! Reflections from that peace of cable caused uneven frequency response in very different places, all over the length of the line.
And do you know how it was fixed?
I fitted a pair of matching transformers at both ends of that cable of the length no more than couple of miles!
In 1981 I moved because of family reasons to a town where jobs for electronic engineers in development area were not available because there were no electronics developments in the area, so I had to accept a position in technical support of railway communications and electronics.
However, it was boring, especially to fix things that existed long before I come there, especially when no modern resources I got used to were available.
One of problems I faced, was a problem with phone ringing on one station.
It was a dispatchers' line, 200 miles long pair of wires on poles, to which many high input/output impedance peaces of audio equipment were connected including special ring receivers that reacted on combination of tones in audio band. Everybody who was connected to the line had to hear each other, but they could be called by sending corresponding sets of tones on which thir equipment must react. That particular station reacted on it's tone set intermittently, and support engineers were endlessly balancing from no rings to false alarms...
All modules, including amplifiers with AGC, filters, demodulators, were replaced and readjusted many times, but it did not help.
When I went there I've found that there is a notch exactly on one frequency of the set officially assigned to call this station, and exactly in that location! The line was properly terminated on both ends. Line taster showed reflections, but on the distance of many miles from that troublesome station. However, it was a good clue, and I've found that as the result of reconstruction on the station in the distance of many miles the line was cut and a peace of underground cable was fitted in! Reflections from that peace of cable caused uneven frequency response in very different places, all over the length of the line.
And do you know how it was fixed?
I fitted a pair of matching transformers at both ends of that cable of the length no more than couple of miles!
Hi Anatoliy,
You had something similar to what we would call a party line. An old telephone type of service.
Even today, we have enough trouble keeping telephone lines matched and balanced, for similar reasons. The average person never thinks about this as they pick up the phone and yap away.
-Chris
You had something similar to what we would call a party line. An old telephone type of service.
Even today, we have enough trouble keeping telephone lines matched and balanced, for similar reasons. The average person never thinks about this as they pick up the phone and yap away.
-Chris
This old argument again?
I've got all the usual objections to the 'subjectivist' position, but I've got a major issue with the typical 'objectivist' view of things.
I don't know of an objectivist who chooses his own equipment objectively. You can't use measurements because the measurements must be interpreted. That's the exact definition of subjectivity!
If you've got two squiggly lines and an ideal measurement is a straight line, which squiggly line is better? Smallest maximum deviation from flat? Smallest area under the curve? And you say: "That's easy. If all are good enough then one is as good as any other." What about two entirely different measurements? Which one is more important?
In order to use measurements as a means of choosing components and call it "objective", is it not essential that given two different sets of measurements it can be established which one most people will prefer in a blind listening test? If you say this can be done, please elaborate. I expect to see quantification of overall perceived sound quality into a single figure of merit. Earl Geddes has done some work in this direction regarding distortion (throwing out THD for something that actually correlates to real preferences) but I'm talking about a figure of merit that combines it all. Frequency response, harmonic distortion and everything else. Not likely!
If you aren't doing large scale, blind listening tests that are independently conducted, you are not behaving as an objectivist no matter how many measurements you make. In some ways many 'objectivists' are worse than 'subjectivists' because they have biases they refuse to see.
I've got all the usual objections to the 'subjectivist' position, but I've got a major issue with the typical 'objectivist' view of things.
I don't know of an objectivist who chooses his own equipment objectively. You can't use measurements because the measurements must be interpreted. That's the exact definition of subjectivity!
If you've got two squiggly lines and an ideal measurement is a straight line, which squiggly line is better? Smallest maximum deviation from flat? Smallest area under the curve? And you say: "That's easy. If all are good enough then one is as good as any other." What about two entirely different measurements? Which one is more important?
In order to use measurements as a means of choosing components and call it "objective", is it not essential that given two different sets of measurements it can be established which one most people will prefer in a blind listening test? If you say this can be done, please elaborate. I expect to see quantification of overall perceived sound quality into a single figure of merit. Earl Geddes has done some work in this direction regarding distortion (throwing out THD for something that actually correlates to real preferences) but I'm talking about a figure of merit that combines it all. Frequency response, harmonic distortion and everything else. Not likely!
If you aren't doing large scale, blind listening tests that are independently conducted, you are not behaving as an objectivist no matter how many measurements you make. In some ways many 'objectivists' are worse than 'subjectivists' because they have biases they refuse to see.
anatech said:
Sure!
And what complicated even more, there were shortwave transceivers connected to the same line in different places, so dispatchers could call machinists in locomotives, and machinists could call dispatchers. Radio transceivers acted like modern cellphone stations, and something like Ethernet protocol was used to connect only one nearest radio transceiver to the line. Fixing the problem with that cable reflections I've found that radio transceivers started behaving more accurately.
But if was the line of 200 miles distance! How long must be an audiophile cable to act similarly on even 20 KHz frequency?
jeff mai said:
Today I was, instead of doing double blind testing, using very biased opinion that in order to get more of undistorted power I need to have more of available current in output MOSFETs, but the current have to be limited so no damage happens anymore when tubes turn on, but gyrator is still slowly reacting. If it would react faster MOSFETs would be safer, but frequency response on lows would be rolled off. Also, resistors in sources could be made of bigger resistance, but it would require higher swing from output tubes and of phase splitter that is limited by low B+. So, trying to find balance of many variables I was soldering, calculating, measuring, resoldering, listening... I did not want to complicate things adding one more B+ source for preamp and phase splitter, but tomorrow will add and continue calculating, measuring, listening, soldering... With absolutely no double blind tests building a peace of equipment that will cause sphincters to relax when people will listen to the music... But now it is a mess of parts and wires, far from it's future excellence, and nothing yet can be blindly-tested, in comfortable chairs, wearing slippers, smocking expensive cigars...
Can you ever imagine design of a totally new peace of equipment with no measurements, but with many double-blind tests?
anatech said:
Draw solace from the fact it takes one to know one. 😉 That little 'sub-harmonizer' was lost in the memory hole until today. In one way audio was more fun back then, if in hindsight a bit unhinged.
Wavebourn said:
Absolutely not! Measurements are essential. I still stand by my previous post, however.
rdf said:
Do you mean a bass guitar effect, like a digital pitch shift, or a frequency divider using trigger?
jeff mai said:
That's why designers are so skeptical to double blind tests: because they can explain source of any type of coloration caused by gear they design, instead of just concluding that "Something sounds different".
Anatech,
All I can find in the archives relating to the 12AT7 tubes tested and sample size is the following:
If you could provide a link to the putative dataset SY referred to I would be eternally grateful.
Now on to the other bit:
If I was both egotistical and silly enough to make a rash, unqualified statement as to the absolute superiority of one tube variant over all others ever produced, based on a limited sample, I would take up your offer and justify my position. In fact I would be obliged to. However, I have not, and would not
The equipment used and accuracy of measurements used are an irrelevance, even if you assume that THD and distortion spectra are the be all and end all of evaluative criteria. If you don't test enough other tubes, and use a sample size large enough to give you a statistically meaningful result you CANNOT make a definitive and unqualifed statement as to the ultimate superiority of one variant . Period.
As I said previously: I’ll bet $10k to James Randi’s left testicle that SY wouldn’t be game to submit a paper containing his dataset and conclusions to a refereed journal. So why publish here? The problem is, some people (especially newbies) actually believe this stuff.
How about: I tested X number of JJ 12AT7 on an HP 3581A tuned frequency wave analyzer, and a (whatever else he used) and made THD and HD spectra measurements. I then compared these measurements with Y samples of a YY 12AT7, Z samples of a ZZ 12AT7, and so on....
The results of these measurements indicated that the JJ produced the least THD of any variant (X%, which was Y% lower than the nearest contender, the ). The JJ spectral analysis showed a nice waterfall effect with diminishing high order artefacts. The YY exhibited high levels of odd order distortion, while the ZZ displayed fairly similar levels of all harmonics up to the 7th, where distortion started to fall, etc, etc.
Not too difficult, accurate, and easy to substantiate. As stated previously, have a look at George's post #450 for a good example of how to present experimental findings in a more considered and accurate manner.
All I can find in the archives relating to the 12AT7 tubes tested and sample size is the following:
If you could provide a link to the putative dataset SY referred to I would be eternally grateful.
Now on to the other bit:
If I was both egotistical and silly enough to make a rash, unqualified statement as to the absolute superiority of one tube variant over all others ever produced, based on a limited sample, I would take up your offer and justify my position. In fact I would be obliged to. However, I have not, and would not
The equipment used and accuracy of measurements used are an irrelevance, even if you assume that THD and distortion spectra are the be all and end all of evaluative criteria. If you don't test enough other tubes, and use a sample size large enough to give you a statistically meaningful result you CANNOT make a definitive and unqualifed statement as to the ultimate superiority of one variant . Period.
As I said previously: I’ll bet $10k to James Randi’s left testicle that SY wouldn’t be game to submit a paper containing his dataset and conclusions to a refereed journal. So why publish here? The problem is, some people (especially newbies) actually believe this stuff.
How about: I tested X number of JJ 12AT7 on an HP 3581A tuned frequency wave analyzer, and a (whatever else he used) and made THD and HD spectra measurements. I then compared these measurements with Y samples of a YY 12AT7, Z samples of a ZZ 12AT7, and so on....
The results of these measurements indicated that the JJ produced the least THD of any variant (X%, which was Y% lower than the nearest contender, the ). The JJ spectral analysis showed a nice waterfall effect with diminishing high order artefacts. The YY exhibited high levels of odd order distortion, while the ZZ displayed fairly similar levels of all harmonics up to the 7th, where distortion started to fall, etc, etc.
Not too difficult, accurate, and easy to substantiate. As stated previously, have a look at George's post #450 for a good example of how to present experimental findings in a more considered and accurate manner.
They are different tubes.
JJ appears to be usable from lower currents.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
JJ appears to be usable from lower currents.
Fixed.
mach1, you might want to practice the search function a bit more. One suggestion- use the "display results as posts" rather than the default "display results as threads."
- Status
- Not open for further replies.