I think that's a pretty important part that factors in to overall amp/speaker performance. Comparisons can be simplified to 2 amp types and 2 speaker types:
Amp A: Tube
Amp B: SS
Speaker X: Low efficiency, high mass cone, low Q
Speaker Y: High efficiency, low mass cone, high Q
Possible combinations are:
AX
AY
BX
BY
I would speculate only one speaker type has been used for the test. It would be interesting if I were mistaken.
According to his web site he (1) has a variety of speakers available, and (2) the subject may bring their own speakers if they prefer.
He does require thet they use dynamic drivers (I presume this rules out electrostatic speakers).
And I beleive he requires the same speakers be used by both amps.
He does require thet they use dynamic drivers (I presume this rules out electrostatic speakers).
And I beleive he requires the same speakers be used by both amps.
If you read the stipulations of the Richard Clark test carefully, you will see that equalisation is applied if necessary, in order to match the frequency responses of the two amplifiers in the audible range. In the event that the tube amplifier had a sufficiently flat response that compensation was not needed, then it would not be used.
Chris
And he started to allow car speakers after 2005 or thereabouts. In the pdf gk7 linked earlier, the speakers used were not reported. It would be interesting to know if Clark has also demonstrated no difference in amps for speakers on different ends of the spectrum of driving requirements. Until then, "all the same" is limited in scope.
It would be interesting to know if the mid boost knob labelled "tube sound" on the Buttler unit was upstream or downstream from the distorting tube. If it's upstream, it might make sense that they decided that it sounds better to distort the midrange frequencies first. So it's a distortion control first, and a frequency selective one second, if that makes sense. The Clapton "signature stratocaster" guitar use to have a mid-boost knob on it, active battery driven, with allegedly 20dB of boost centered around 800HZ according to his sound man (Dickenson). Maybe that's what a guitarist prefers when they are going to solo. Get rid of the delicate strat sound and get manly...
Subject to the basic stipulations listed in the challenge, the contender is free to choose the amplifier, speakers, etc. in whatever way they feel will maximise their chances of winning the prize. Of course only the amplifier is swapped in the A and B double-blind testing, but the challenger can specify the choice of speakers that will be used in the test, within the stipulated limits.
Chris
The pdf I posted is from an ABX test Stereophile conducted, it´s not related to the "Clark challenge".
Ah alright then. Are the speakers listed on another page for that article?
cnpope said:
I understood the rules before posting. The question is whether the EQ/output impedance adjustments are effective for all speaker/amp interactions or if these adjustments are just effective for a subset of speaker types.
My testing with an oscilloscope supports what Bob Richards says. I fitted a resistor across the output of the amplifier, audio generator and ran the amplifier at full power and a little more so that the sine wave showed the flat tops you get a little above maximum power. The difference can be seen in how sharp the junction between the flat distorted bit and the sine wave is. The transistor amplifier had sharp corners at the point of distortion setting in. The tube amplifier had rounded corners at this point. So when the transistor amplifier is over driven the transistors hit the limit where signal meets the power supply voltage very hard. Tubes have a much softer landing at this point with rounded ends to the flat bit. At lower power levels there is very little difference between the two amplifiers if any at all. My conclusion is that the tube sound can be had when the amplifier is driven close to the limit. No wonder guitar owners like tube amplifiers.
Last edited:
According to that, there was one person (believer) with a 94.4% probability of being able to distinguish between amps in an ABX test and one skeptic with an 84.4% probability of being able to distinguish between the two amps in an ABX test.
(The probability of being correct is 1 - "probability due to chance".)
You are referring to this
and I replied,
You see, Pano, your statement "fidelity to the original sound, not always to the original signal", is incorrect when describing amps. With amps, it is always about the signal. Signal goes in and signal comes out. That's what amps handle.Except for the loudspeakers, what other ways do electronic audio components work besides sending electronic signals?
It helps to use quotes when responding to what you think someone stated.
How many sessions was that test done by that person and what was the setup condition like? Has he been able to repeat that in other ABX?
Note that this was a test of the Counterpoint amp, which blew up during listening. They were notorious for being unstable, and that could certainly be audible. The pdf gives the number of trials, if that's what you mean by "sessions."
The test of Futterman vs Pioneer also shows a significant probability of being able to discern a difference (84%).
That's actually not particularly a good number (and it's not quite how you defined it)- generally, a 95% confidence is required for something to be thought of as probably significant.
Nonetheless, I wouldn't be surprised if in a test of a few more people, you could hit a 95% confidence with the Futterman, since it has a relatively high output impedance.
I agree that the numbers do not give strong confidence in the results.
That said, they are interesting in that they may present some results that are counter to the tests Clark did.
Then again, if he is correct, they may not have adjusted the amplifiers within the required matching for levels and that according to Clark would have been enough for the results seen.
In addition, I believe Clark would have adjusted the output impedance of the Pioneer to match the Futterman which might effect the results as well.
That said, they are interesting in that they may present some results that are counter to the tests Clark did.
Then again, if he is correct, they may not have adjusted the amplifiers within the required matching for levels and that according to Clark would have been enough for the results seen.
In addition, I believe Clark would have adjusted the output impedance of the Pioneer to match the Futterman which might effect the results as well.
Exactly so. Into a real world load, changes in amplifier output impedance cause changes in frequency response, something to which humans are demonstrably sensitive. That can sometimes be simmed with a resistor, but in some cases (high output impedance that varies with frequency), you have to get a bit fancier with cheap Rs and Cs to get a good match.
IIRC, the distortion components didn't change at all. IOW, the knob did absolutely nothing except for control the level of midrange boost. (Prof. Ken Pohlmann bench-tested the amp in question for a magazine.)
Also, this boost was centered about 2 octaves up from the boost in the Clapton amp you mention.
- Status
- Not open for further replies.