|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
82.1.230.24
In Reply to: RE: Listen for yourself posted by Donald North on July 07, 2017 at 09:13:39
"I heard my first SET amplifier (an Audio Note UK system) back in 1994 when I was 21. It had more instrument texture, nuance, natural clarity and presence than anything I heard heard to date, and while an undergrad at Caltech I spent all of my free time visiting every stereo store I could in LA to hear what was available."
I don't think anything I said would be in contradiction with what you are saying. I am merely saying that I think there are rational scientific explanations for why this would be the case.
Chris
Follow Ups:
The most common "explanation" for the SET sound is gentle clipping and 2nd harmonic distortion. I have made SET phono stages and headphone amplifiers, both of which are operated no where near clipping and have easy loads to drive, which have very low measured distortion yet sound noticeably different from other devices, be they solid state or tube. I believe there is more to the picture, yet to be discovered and understood, than standard bench tests.
Perhaps I should expand upon my previous response to your posting.I apologise if I picked the wrong word by describing the sound from an SET amplifier as "pleasing," since it might perhaps have carried a connotation of condescension that was not intended. I was just looking for a word that conveyed the idea that the sound gave a more satisfying all-round experience for the listener. And in that sense, I would describe a system that "had more instrument texture, nuance, natural clarity and presence than anything I heard heard to date" as a system that had a more "pleasing" sound. Perhaps there is a better word I could have used.
Now, as regards my statement that I suspect that these phenomena can be accounted for in terms of rationally understandable scientific principles, let me expand a little on this. Of course, ultimately, one (or at least a scientist such as myself) would wish to be able to explain everything we observe in terms of the most fundamental concepts and building blocks (quarks, leptons, strings,...????), but obviously at present that is not feasible. But explanations of phenomena in terms of more general observed phenomena can also be perfectly "scientific," even if they do not go all the way back to the level of the fundamental building blocks.
In the present context, I would take the observed psycho-acoustical characteristics of the human mind and ear as scientific phenomenological observations. A couple of examples are:
1) Humans find mixtures of frequencies that are related by simple rational ratios to be pleasing to the ear, and on the other hand irrationally-related mixed frequencies tend to sound unpleasant. Based on this observation, one can predict that if one made a sound-producing system in which all frequencies were shifted upwards or downwards by a constant additive amount, then music played through this system would sound absolutely dreadful. And this is, I believe, borne out by experimental observations. (A simple way to test this is if one has a shortwave receiver with narrow-band filtering that can block the central carrier frequency, and one then replaces this with a local BFO operating at a slightly different frequency.)
2) Humans find that the addition of the second harmonic to a tone or set of tones can provide a pleasing (in my generic sense) alteration to the overall effect. In particular, in a piece of music it can lead to an augmentation of the feeling of "liveliness," and maybe it can allow the brain to interpolate or extrapolate and give more of a lifelike overall experience. (Research in psycho-acoustical phenomena, as with many other aspects of the interaction of the brain with our exterior senses, seems to indicate that the brain does an amazing job of "filling in" the things that are inadequately conveyed to it by our sense organs.)
As a rather trivial hypothetical example, if it were the case that broadcast music stations used single or double-sideband AM transmissions where the carrier was removed and needed to be restored in the end-user's receiver, then one could easily make a general "scientific" observation that the best-sounding receivers would be those that restored the correct suppressed carrier frequency, rather than one that was displaced up or down by some amount. This would not be in the least bit controversial.
In a similar vein, it is a perfectly legitimate scientific endeavour to look for what characteristics of an SET amplifier might be responsible for giving humans the feeling that the performance is more "alive," or have more instrument texture, nuance and clarity. It seems reasonable to suspect that the phenomenological observations noted in point (2) above could be playing an important role, especially when one notes that one of the most striking distinguishing features of an SET amplifier is the relatively large percentage of second-order harmonic distortion.
Thus, I don't see why there should be anything controversial about the suggestion that if one can identify certain generic and measurable characteristics of SET amplifiers that distinguish them from most other amplifiers, then it is quite likely that it is these characteristics that are playing an important role in accounting for why they produce the audible sensations ("presence," "nuance," "instrument texture," or whatever) that they do. Thus, technical measurements of an amplifier's distortion and other characteristics (such as frequency response) might very well allow one to predict what kinds of overall impression it could give to the listener. You may be right that there are further characteristics of an SET amplifier that have yet to be identified and measured, but I wonder if there is really solid evidence for this?
Chris
Edits: 07/07/17
You say again "one of the most striking distinguishing features of an SET amplifier is the relatively large percentage of second-order harmonic distortion." As I previously said, I have made SET phono stages and headphone amplifiers which have very low measured distortion in their normal operating range yet they still sound different and I believe superior. I don't believe this is attributable to 2nd harmonic distortion when the THD is less than 0.1%. Another differentiating factor which you have not mentioned is the fact that they do not split or mirror the phase as do differential and push-pull amplifiers do. I believe this is a contributing factor to the SET sound.
An example: Back in the 90s Audio Note UK made two tube amplifiers that were as identical as possible, one a parallel single ended, the other a push-pull. The reviews all said the single ended version sounded better.
"I have made SET phono stages and headphone amplifiers which have very low measured distortion in their normal operating range yet they still sound different and I believe superior. I don't believe this is attributable to 2nd harmonic distortion when the THD is less than 0.1%"
I think simple SE amplifiers made with direct heated triode output tubes that do not use feedback are notable by what is lacking in the HD. Namely, upper order HD.
I believe it's not the (relatively) high amounts of 2nd HD that make them "sound so good", it's the lack of 5th, 7th and 9th.
I believe (but can not prove) that the smallest amount (maybe even an amount that can't be measured) of 7th and it's game over. The electronics sound like electronics instead of music.
I could be proven wrong on all counts but this is what I believe and it is what guides me when designing circuits and picking operating points.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
and Crowhurst even fought for a weighting system which would have assigned higher penalties for 5,7, and 9. There is text somewhere on the net about him and some BBC guys advocating for this but were shouted down by manufacturers. and they specifically mentioned upper odd harmonics at some crazy low levels being very annoying.
Yes and these were serious proposals.One would be multiplying the percentage measured by the order. 5th would be the measured % times 5. 7th would be the measured % times 7.
The other proposed method would have been the % measured times the square of the order. 5th would be the measured % times 25 and 7th would be the measured % times 49.
These proposals were intended to better reflect the annoyance factor.
THD doesn't tell us much in terms of the annoyance factor. We need to now the orders of the HD.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 07/07/17
"and Crowhurst even fought for a weighting system which would have assigned higher penalties for 5,7, and 9. There is text somewhere on the net about him and some BBC guys advocating for this but were shouted down by manufacturers. and they specifically mentioned upper odd harmonics at some crazy low levels being very annoying."I agree that that is a very interesting point, and maybe is highly relevant. Do you, as a matter of interest, know if SET amplifiers tend to have lower odd-harmonic distortions than PP amplifiers? I just don't know one way or the other; I don't recall ever having seen studies of this question.
In any case, such things should lie within the threshold of measurability, presumably, and so they would be part of the whole package of "measurements that can be correlated with observation" in the spirit of the scientific method.
Chris
Edits: 07/07/17
"I believe it's not the (relatively) high amounts of 2nd HD that make them "sound so good", it's the lack of 5th, 7th and 9th.I believe (but can not prove) that the smallest amount (maybe even an amount that can't be measured) of 7th and it's game over. The electronics sound like electronics instead of music."
You may be right that it is a lack of higher odd-harmonic distortions that is responsible. (Is this actually a property of an SET, by the way? I can see why it would tend to have more even-harmonic distortion than a PP amplifier, but is there a reason why it would have less higher odd-harmonic distortion? I'm not doubting, just asking.)
However, what about the suggestion that the higher odd-harmonics below the threshold of measurability, could be responsible? A modern state-of-the-art spectrum analyser is amazingly sensitive. Is there really any reason to suspect that amounts below the threshold could have audible effects? Or is this an example of the very human desire (in some humans at least) to believe that there are things that lie beyond our understanding? If it could be demonstrated in convincing experiments that such effects were occurring, I would be only too happy to accept them. But I wonder what evidence there might be?
As a matter of practicality, I would have thought that any real-world amplifier in existence today would have 5, 7 and 9th order distortions that are actually within the range of measurability by present-day spectrum analysers, and so speculation about the possibility of such distortions below the measurable threshold would be, at this stage at least, academic. (Again, if I am wrong about this, I would be happy to be corrected.)
Chris
Edits: 07/07/17
Chris
It has been noted that the use of negative feedback can reduce second harmonic distortion, however Dr. Earl Geddes (among others) has noted that added second harmonic distortion is relatively difficult to hear compared to other types of distortion. The second harmonic is an octave above the fundamental and harmonizes with the fundamental which makes it difficult to separate in our perception from the fundamental. While negative feedback reduces the second harmonic, it can actually increase higher odd order harmonic distortion. As the second harmonic is the biggest number, reducing it makes the total harmonic distortion number look better at the expense of increasing higher order odd harmonic content which has been shown to be perceived as unmusical. But you don't have to believe me on this, Jean Hiraga wrote several articles examining this subject going back into the late 70's and 80's which we didn't get here in the US unfortunately until the oughties when they were reprinted in AudioXpress. See the link here for the Amplifier Musicality article which was followed by rebuttals from two experts, and which was followed by a rebuttal by Hiraga with no rebuttal to Hiraga's rebuttal.
Paul
I'm pressed for time right now but a DHT is very linear. If loaded and driven correctly, and without NFB, there is not much of a mechanism for producing upper ordered HD."Norman Crowhurst wrote a fascinating analysis of feedback multiplying the order of harmonics, which has been reprinted in Glass Audio, Vol 7-6, pp. 20 through 30. Mr. Crowhurst starts with one tube generating only 2nd harmonic, adds a second tube in series (resulting in 2nd, 3rd, and 4th), and then makes the whole thing push-pull (resulting in 3rd, 5th, 7th, and 9th), and last but not least, adds feedback to the circuit, which creates a series of harmonics out to the 81st. All of this complexity arises from theoretically-perfect tubes that only create pure 2nd harmonics!"
__
I believe there are studies showing that humans can hear well down into the noise floor.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 07/07/17 07/08/17
I think you may have to look closely at the word rational. Let's say we know 100 things about audio we can test. Could there be 100 more we do not know yet. Infinity is a pretty big place. To go from I know a few things about the human body and energy to I know everything is hardly what I would
call rational.
So the real question is who is drinking the koolaide???
Enjoy the ride
Tom
" Let's say we know 100 things about audio we can test. Could there be 100 more we do not know yet. Infinity is a pretty big place. To go from I know a few things about the human body and energy to I know everything is hardly what I would call rational."
Well a question one has to ask is, is there any compelling evidence that the 100 things we already know are insufficient to account for the observed phenomena? I get the impression sometimes that some people just don't want the things we already understand to be able to explain the things we observe.
As I have said before, I am not for one moment disputing that SET amplifiers may sound different, and indeed better, than other amplifiers. There are some "obvious" reasons why this might be the case. Is there compelling evidence that the obvious and already understood reasons are insufficient to explain the observed phenomenon?
Chris
Why are there no commercial applications for SET amps? What amps were used when the recording mixer did his work?
Wait a minute: The analog record circuitry in the classic Ampex 351 tube tape recorder was all single ended. Its internal playback electronics were push-pull.
Edits: 07/07/17
Of course lots of 1950/60s studio and broadcast line level gear was single ended and it worked just fine.
Gusser, I'm curious to know, though, what kind of amplification and processing stages a typical audio signal will have gone through, in a modern recording studio and CD production plant, on its way from the studio microphones to the output of the home-user's CD player.
Would the amplification stages typically be op-amps, with lots of feedback? Could you walk us through the steps the signal might typically follow, between the microphones and the CD player's output? That would be really interesting.
Chris
Today most work is done on digital workstations. The large mixing consoles with multiple flat screens are just control surfaces talking to a bunch of commodity PC's over Ethernet. The audio processing is all via AES in and out - so much for all the SPDIF jiter scares! AES and SPDIF are virtually identical - especially in terms of jitter performance.Now the microphone preamp before the ADC is analog of course and mostly an OPAMP design. There are some studios that use discrete transistor mic preamps and some even tube preamps. But beyond the ADC, the audio remain digital, in fact a computer file. Tape is long gone except again for tiny esoteric operations. All storage is via hard disk and that data is shipped and modified around a lot including between facilities over the public internet - encrypted of course.
In the 1980s, large mixing consoles were all OPAMP. Hundreds of them in a 128 input board. Anywhere from two to ten stages per module. I once cared for a Neeve console that had 128 inputs and the internal bussing was wide standard ribbon cable. They ran balanced +/-/G across a 50 wire cable with no twists either. No crosstalk and no noise. It just shows what is possible with god engineering as Neeve is well known for. So much for esoteric audiophile cables! If any of that hype was true, Neeve would have used them. Cost was no object on his products.
The 1970 was the rein of discrete transistor OPAMPS. And they were not as good to spite what some audio magazines want you to believe.
Edits: 07/10/17 07/10/17
Thanks for the summary of the path from the microphone; very informative.
Presumably there is quite a bit of analogue processing after the DAC in the home CD player? Low-pass filtering to block the digitisation artifacts, and so on? Op-amps, typically?
What I'm thinking, then, is that probably at the two ends of the chain (microphone amplifier, and output stages of the CD player) the typical setup will involve op-amps, with the usual very large negative feedback applied. I'm just wondering where that leaves the people who argue that negative feedback is a bad thing.
Chris
After listening to a fair group of SE amps, I have yet to hear one I wanted to build, except for geetar amps.
For stuff SE needs, like gapped OPT's there is some benefit to the linearizing effect this has on the sonics delivery. I may try gapped PP at some point if I can live with the reduction in LF performance the drop in overall primary L gives. Since I run electronic cross overs, this could be sooner than later...
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Bi-amping is your friend: I run parallel single ended 2A3s above 160Hz and McIntosh push-pull tubes in the bass.
"Bi-amping is your friend"
I totally agree.
Gapped transformers and good bass performance are, IMO, mutually exclusive.
The gap reduces the inductance, the reduced inductance causes the loadline to become elliptical (in the presents of low bass) and all the frequencies follow that elliptical loadline increasing the HD.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Yah, it is a fight...for the bass amps, I think I want as much as I can get. Anything else above that, the gapped core sonics are not to be despised.
It does not take anything like a SE, full Class A idle current worth of a gap to be of use. Just enough gap to knock 10% off is quite worthwhile. it is likely one of the reasons I like cut C-cores; their effective gap is much larger than an interleaved stack of E's and I's.
I think I need some 49% Ni cut C's...
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"Why are there no commercial applications for SET amps? What amps were used when the recording mixer did his work?"
Indeed, this gets to the nub of the question. When someone chooses to use either a tube SET or PP amplifier, or an SS amplifier, in their home stereo system, they are just making a selection that sits at the end of a long chain of totally "conventional" SS amplification and processing stages. If this last step of the chain makes a big difference, it seems likely that it is because of what it is adding, rather than what it is not adding, to the signal that has been fed into it.
I have never understood why this is a controversial point, but somehow the suggestion that the SET amplifier is adding colourations that are "pleasing," "nuanced," or whatever, seems not to go down too well sometimes!
Chris
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: