tvi said:
Thanks a lot for posting this link James. Never knew there was a patent. Looks quite entertaining.
I would like to point out, for clarity, that there are 3 major contributions to making a successful sounding amplifiier design. I say 'successful sounding' because there are other factors that might be just as important in a successful product, such as looks, reliability, and features. I want to concentrate on the 'sound quality' of the product here.
That is composed of: topology, layout, and execution.
It took me years, but I came to realize that I am only really good at one of these areas. That is circuit topology. Back in the '60's, I developed most of the circuit topologies that I use today, pretty much by myself, but I did look closely at the work of others, such as Bob Widler. There I saw the freedom to experiment and to make DC coupled designs, when at the time, audio was virtually all AC cap coupled, with single ended power supplies.
In the '70's, I used to have others lay out my designs or I made them on vector board. For a few years, I had the help of Swiss trained techs, who could really make a prototype, that worked well and sounded ok.
Then, about 30 years ago, I partnered up with an Englishman, to create my own company and he did the circuit layouts. They didn't look too good, but that was not my area, so I stayed quiet, but when one battery powered pre-preamp circuit, composed of only 2 transistors per channel, HUMMED when put close to an AC field, even with a thick case around it, I became concerned. Something was wrong here, and when UC Berkeley, overtly rejected his layout of a design I was helping them with, the nail was in the coffin. Yet, I could NOT tell him of his problem. He just would not believe it. I must admit that he was the best model chassis maker that I have ever worked with, he was an expert with bending metal, but we went our separate ways.
Then, when I came up with the Vendetta topology, I again needed a 'layout' man and I hired a very nice guy who did a pretty good job, and we got started, but then he moved on to Dolby, where he is today, with my blessing, as I could not afford to support him and he had a family.
Almost by chance, I met Carl Thompson, who I still work with, today. He was magic! His layouts almost glow in the dark! Wow! And Vendetta Research was launched. Those original Vendetta boards are about as good as anything that we can make today, sonically, and are still used today by selected people around the world. I even use them in my CTC phono stage.
It takes: Heart, skill, and imagination to make a good circuit layout, and many here are not up to the task, including me. It is best to realize this, and find someone who can do it best.
The final area is 'execution'. Most here would laugh at what we consider important, such as which brand and type of connector, which connecting wire, solder, and dozens of other things. However, this is how we get A ratings and still keep our prices within some reasonable cost. Without 'execution' we fall right off the rating scale with 'Stereophile' and most audiophiles. We have tried it often enough, to be sure. This was the responsibilty of my old colleague, Bob Crump. He could easily hear differences, and I took his advice to heart.
Unfortunately, now I am on my own. Bob is gone, but his ideas, hopefully, will stay with me for the next generation of designs.
That is composed of: topology, layout, and execution.
It took me years, but I came to realize that I am only really good at one of these areas. That is circuit topology. Back in the '60's, I developed most of the circuit topologies that I use today, pretty much by myself, but I did look closely at the work of others, such as Bob Widler. There I saw the freedom to experiment and to make DC coupled designs, when at the time, audio was virtually all AC cap coupled, with single ended power supplies.
In the '70's, I used to have others lay out my designs or I made them on vector board. For a few years, I had the help of Swiss trained techs, who could really make a prototype, that worked well and sounded ok.
Then, about 30 years ago, I partnered up with an Englishman, to create my own company and he did the circuit layouts. They didn't look too good, but that was not my area, so I stayed quiet, but when one battery powered pre-preamp circuit, composed of only 2 transistors per channel, HUMMED when put close to an AC field, even with a thick case around it, I became concerned. Something was wrong here, and when UC Berkeley, overtly rejected his layout of a design I was helping them with, the nail was in the coffin. Yet, I could NOT tell him of his problem. He just would not believe it. I must admit that he was the best model chassis maker that I have ever worked with, he was an expert with bending metal, but we went our separate ways.
Then, when I came up with the Vendetta topology, I again needed a 'layout' man and I hired a very nice guy who did a pretty good job, and we got started, but then he moved on to Dolby, where he is today, with my blessing, as I could not afford to support him and he had a family.
Almost by chance, I met Carl Thompson, who I still work with, today. He was magic! His layouts almost glow in the dark! Wow! And Vendetta Research was launched. Those original Vendetta boards are about as good as anything that we can make today, sonically, and are still used today by selected people around the world. I even use them in my CTC phono stage.
It takes: Heart, skill, and imagination to make a good circuit layout, and many here are not up to the task, including me. It is best to realize this, and find someone who can do it best.
The final area is 'execution'. Most here would laugh at what we consider important, such as which brand and type of connector, which connecting wire, solder, and dozens of other things. However, this is how we get A ratings and still keep our prices within some reasonable cost. Without 'execution' we fall right off the rating scale with 'Stereophile' and most audiophiles. We have tried it often enough, to be sure. This was the responsibilty of my old colleague, Bob Crump. He could easily hear differences, and I took his advice to heart.
Unfortunately, now I am on my own. Bob is gone, but his ideas, hopefully, will stay with me for the next generation of designs.
I would further like to point out something that I have learned, that many engineers just don't want to accept.
Audio design often involves knowledge, mystery (we don't know everything yet), and sometimes, luck.
Think of the parable of the doctor who bragged about the operation that he just performed, even though the patient died. This is how some engineers design equipment. They think that they have done everything right, and they certainly know what is right, because they went to engineering school, and if people don't like the sound of what they designed, it is THEIR problem, not the designer's. 😉
Audio design often involves knowledge, mystery (we don't know everything yet), and sometimes, luck.
Think of the parable of the doctor who bragged about the operation that he just performed, even though the patient died. This is how some engineers design equipment. They think that they have done everything right, and they certainly know what is right, because they went to engineering school, and if people don't like the sound of what they designed, it is THEIR problem, not the designer's. 😉
It is not just that. It is the understanding that formal knowledge is not experience, and you need both. Just like doctors out of med. school. At least, they make them work with experienced doctors, before they are let loose on society. 😉
Hi John,
I don't think you will have any arguments from any professional in any field there. This is a condition that crosses every area of study.
Sometimes pride can be blind. It also gives rise to our favorite "not invented here" issue. That's where people ignore any advance or different way of thinking not developed in their direct community.
-Chris
I don't think you will have any arguments from any professional in any field there. This is a condition that crosses every area of study.
Sometimes pride can be blind. It also gives rise to our favorite "not invented here" issue. That's where people ignore any advance or different way of thinking not developed in their direct community.
-Chris
john curl said:
john curl said:
John, great job on the JC-2, and the cost is really quite reasonable. I agree with you on John Atkinson's measurements for Stereophile, and the new AP has even better resolution. His stated goal is to try to correlate the objective with the subjective - a major task that may or may not be possible even with state-of-the-art test equipment. All the other audio magazines treat technical guys like children...."there, there. You really don't have any need to see all that confusing and complicated data. Here, we have some new adjectives for you to play with"
Best Regards, Chuck Hansen
Until someone does find a way to correlate measurements with the sound quality, what purpose does it serve to repeat numbers that are meaningless (and can readily be found on the manufacturer's website)? To impress people? Who? The only people impressed by the standard numbers are people who haven't listened to two amps which on paper have identical specs, yet sound notably different. People who have listened are justifiably skeptical of THD, damping factor, etc. There's no reason to trot out the numbers except to give a thin veneer of respectability to the review.
The only exception as I see it would be a component that did not meet specifications. That, in turn, raises the question of whether the piece was damaged in shipment or was perhaps defective...or did the manufacturer claim more than they could deliver? In this case, I believe that it would be interesting, not so much from the direct implications of the numbers for sound quality, but for what you might deduce about the manufacturer's honesty.
All of this could change in a heartbeat if someone comes up with a new spec that actually means something, but I'm not holding my breath. Progress is slow to say the least. In the meantime, save the ink, save the paper.
To the extent that someone claims to be trying to find a correlation, my hat's off to them. But are they really? Where's the article detailing their research, whether the results are positive or negative? I, for one, would be interested to read an article like: Well, we thought we'd see if damping factor actually correlated to perceived tightness of bass and here's what we found. But oddly, I never see articles like that. On the rare occasion someone does attempt something of the sort, they only compare two or three amps, which is hardly going to give you a statistically worthwhile answer; how do you know the bass wasn't better due to a bigger power supply, for instance--something unrelated to damping factor. In short, I think a strong case could be made that (other than catching a few components that don't meet spec) the reviews in question are simply printing the charts and numbers to make their reviews more 'scientific' in some peoples' eyes. If I want the specs, I can always look at the literature for the piece.
I hope someone, somewhere comes up with something clever. It would be nice to have a reliable test that actually told us something about sound quality.
Grey
The only exception as I see it would be a component that did not meet specifications. That, in turn, raises the question of whether the piece was damaged in shipment or was perhaps defective...or did the manufacturer claim more than they could deliver? In this case, I believe that it would be interesting, not so much from the direct implications of the numbers for sound quality, but for what you might deduce about the manufacturer's honesty.
All of this could change in a heartbeat if someone comes up with a new spec that actually means something, but I'm not holding my breath. Progress is slow to say the least. In the meantime, save the ink, save the paper.
To the extent that someone claims to be trying to find a correlation, my hat's off to them. But are they really? Where's the article detailing their research, whether the results are positive or negative? I, for one, would be interested to read an article like: Well, we thought we'd see if damping factor actually correlated to perceived tightness of bass and here's what we found. But oddly, I never see articles like that. On the rare occasion someone does attempt something of the sort, they only compare two or three amps, which is hardly going to give you a statistically worthwhile answer; how do you know the bass wasn't better due to a bigger power supply, for instance--something unrelated to damping factor. In short, I think a strong case could be made that (other than catching a few components that don't meet spec) the reviews in question are simply printing the charts and numbers to make their reviews more 'scientific' in some peoples' eyes. If I want the specs, I can always look at the literature for the piece.
I hope someone, somewhere comes up with something clever. It would be nice to have a reliable test that actually told us something about sound quality.
Grey
Numbers do matter, and I am glad that 'Stereophile' still measures them. However, you have to interpret the numbers properly to get anything that is really useful. Most people can't do that, and it is not in any book.
GRollins said:
As many times, you do not know what you speak about. There are no 2 different audio components with same set of measured graphs and spectra. The only problem is that you have to understand those measurements, and to add some not shown. Do not disapprove something you apparently do not understand.
I would say, Pavel, that Grey is entitled to any opinion he wants.
Furthermore, your implied arrogance here is vaulting, and even you, for all your self-professed knowledge, cannot fully correlate measurement with statistically based listener preference. In truth, no one can, not even Earl Geddes - who comes closest in my view.
This obsessive concern for the measurements might one day reveal something very worthwhile, but not yet, and your opinion is not worth more than others here, in part because of your sneering attitude to others. Until this correlation is well understood from every angle - PSpice analysis, distortion spectra, operating points, topology - then good design will remain something of an art and certainly will require multiple iterations to get right.
Hugh
Furthermore, your implied arrogance here is vaulting, and even you, for all your self-professed knowledge, cannot fully correlate measurement with statistically based listener preference. In truth, no one can, not even Earl Geddes - who comes closest in my view.
This obsessive concern for the measurements might one day reveal something very worthwhile, but not yet, and your opinion is not worth more than others here, in part because of your sneering attitude to others. Until this correlation is well understood from every angle - PSpice analysis, distortion spectra, operating points, topology - then good design will remain something of an art and certainly will require multiple iterations to get right.
Hugh
A very impressive preamplifier!
Mr Curl, I saw the JC-2 pre uses an R-core transformer. Since it appears to me these are seldom used and there's little hands-on experience with these, I would like to ask you what your experiences with these type of transformers are (though I know that it was probably not your decision to use them).
Thank you very much, Hannes
Mr Curl, I saw the JC-2 pre uses an R-core transformer. Since it appears to me these are seldom used and there's little hands-on experience with these, I would like to ask you what your experiences with these type of transformers are (though I know that it was probably not your decision to use them).
Thank you very much, Hannes
According to this paper, Moore et al even come closer.AKSA said:
The real "problem" is that virtually nobody in the industry (both the audio test gear and audio reproduction branches thereof) has adopted these new and way better sound quality metrics (and quite complicated to derive), at least to my knowledge. A single number THD value and even the full spectrum (mag/phase) it was derived from is close to meaningless if one doesn't know how to correlate that to the perceived sound quality. That's where Geddes/Lee, Moore etc spent their efforts in: a more useful and more realistic interpretation of the given data. There is still a lot of work ahead, say correlation of measurements of stereo channels to perceived "imaging", for example... it will be done, sooner or later.
- Klaus
It does take more than one measurement to give a meaningful evaluation. However, it is possible to derive a lot of subjective information from the measurements in a 'Stereophile' review, especially if the better test equipment is used. Unfortunately, the older set-up had residuals that implied higher order distortion on many products, at a level that was inaccurate.
- Status
- Not open for further replies.