Tube vs Solid State Is the War

[fas42]

And by the way, I've heard and compared the master tape to a high rez copy thru a darn good system using the Playback designs DAC. Sadly, the digital copy is far from a mirror image of the tape. In fact, it's a poor facsimile in many respects, not the least of which is the upper octaves. Cymbals don't sound like any cymbals that I've heard. And digital just flattens the images not to mention, loses the overtones of the instruments.

Myles, I agree with you about the disappointment of hearing so-so digital. The audio industry seems to have such a hard time really understanding what it needs to do to get good sound happening with digital as a source: no matter how much money and technical expertise they throw at it, it frequently is flat as a tack, as you say. At least part of the problem is that everyone is obsessed with the easily measured performance parameters, and ignores the fact that you have to be very fussy about much less obvious interference aspects of the situation.

With a bit of luck, though, and another 10, 20 years or so of fooling around by all and sundry, better understanding eventually should emerge as to how to make digital always sound right ...

Frank

 

[Gregadd]

Taking this and combining it with the original point of that started this thread, what do we think the measurements show of a tube amp that doesn't sound like SS and one that does? Is there nothing revealing in there?

Amir I think I see where you are going.

Certainly both tubes and transistors have faults. You have to consider a competent design not constrained by cost considerations or the load it is driving. A wicked load can make many a good amp sand bad.

 

[Gregadd]

Hmmm. Jazz Trio or model with wings. That's a tough one.

That picture reminds me a fictional book I read. Someone had successfully used the new found knowledge of genetics to breed humans who could use wings to fly. Unfortunately she escaped and set out to free the considerable number of freakish failures.

 

[Phelonious Ponk]

Reread my post and don't put words in my mouth Tim. I was comparing a master tape with a CD. And by the way, I've heard and compared the master tape to a high rez copy thru a darn good system using the Playback designs DAC. Sadly, the digital copy is far from a mirror image of the tape. In fact, it's a poor facsimile in many respects, not the least of which is the upper octaves. Cymbals don't sound like any cymbals that I've heard. And digital just flattens the images not to mention, loses the overtones of the instruments.

If you think that in reality, CDs have greater dynamic range, then I want what you're smoking. CDs have never come close to their "theoretical" dynamic range for many reasons. And you can look up the evidence as everyone from John Atkinson to recording engineers have published data showing what CDs actually contain. And I've posted a link to this info a while back.

And please spare me on the measurements. As a moderator, you should know why as it's even been recently reposted. But as long as you bring it up, please enlighten us on what measurements you think have any correlation to anything, much less tonality and acoustic space. And please look up and read KOJ's measurement work before answering this question. And as far as distortions go, please own up to each medium has its own, unique set of distortions. To my ears, analog is easier to listen to than digital esp. with complex orchestral music.

And while many say the simpler the music, the better digital performs, that's not the case in my experience. I've heard a live guitar recorded on SACD and DVD audio and then played back immediately. The recorded guitar sounded nothing like the original guitar; it was missing delicacy, overtones, body, rhythm, etc. I'll leave it to you to figure out the measurments; I know what my ears heard.

Ok, I re-read it, and yes it's still about tape, we agree on that. I did spare you the measurements, though you seem to be a bit conflicted on that request as you later left it to me to come up with them. No thanks. I know they're out there, I know what they say, and so do you. You evidently believe that the measurements do not reflect what you hear. I'm fine with that. You hear what you hear. Observation. Perception. Opinion. My only challenge to you was to find a way to speak of what your ears hear in those terms, not in definitive, pseudo-scientific language that makes it sound as if you are talking about established facts supported by data, not your own opinion, when, in fact, the opposite is true.

Unless I'm dead wrong about what is supported by the data, in which case, please, don't spare me the measurements. I'd love to see them.

Tim

 

[Gregadd]

Let me see now, one of the leading proponents of the tubes are better camp concedes that the the gap between tubes and solid is closed. Solid state advocates are unwilling to accept that concession.

Interesting.

 

[DaveyF]

And while many say the simpler the music, the better digital performs, that's not the case in my experience. I've heard a live guitar recorded on SACD and DVD audio and then played back immediately. The recorded guitar sounded nothing like the original guitar; it was missing delicacy, overtones, body, rhythm, etc. I'll leave it to you to figure out the measurments; I know what my ears heard.

Myles, I have done this test with not just digital playback but with analog as well. Unfortunately, there is little difference in the results. The 'live' guitar sounds very different than the reproduced. I have also done this test with tape....slightly( VERY slightly) more convincing, however, still not able to get the reproduction absolutely correct. Nothing, I repeat NOTHING, is able to reproduce the sound of a 'live' guitar so that it sounds like the 'real' event just played, IMHO. If someone thinks this is NOT correct, I would venture to say that that someone just hasn't heard a 'live' guitar played up close or perhaps isn't hearing too well, OTOH I suppose FMMV. ( Frank's mileage may vary)

 

[MylesBAstor]

Taking this and combining it with the original point of that started this thread, what do we think the measurements show of a tube amp that doesn't sound like SS and one that does? Is there nothing revealing in there?

That's a great question Amir and having been a scientist, deep down I always feel that we can figure out a way to measure these differences. I guess, unfortunately, reductionism has been of limited usefullness in many applications esp. complex systems. But considering the number of people that have worked on both sides of the output, and some pretty bright ones a that, no designer had to date come with anything meaningful. OTOH, many designers don't care (now I'm not saying they don't measure their products here) and leave that business up to the academic, measurement people. And certainly someone such as Vlad of LAMM is esp. proud of his product's measurements and post them all on his website. But ss as well as tube designers willingly acknowledge the other technologies viewpoint

Point is that there's lot of technical reasons why tubes should sound inferior but it's like the why a bumblebee can't fly argument.

 

[Ron Party]

Taking this and combining it with the original point of that started this thread, what do we think the measurements show of a tube amp that doesn't sound like SS and one that does? Is there nothing revealing in there?

Very profound questions, neither of which seemingly has been or will be answered by those on their own soap boxes who diligently advocating the superiority of one over the other, science be damned. Golden ears trump science.

 

[MylesBAstor]

Myles, I have done this test with not just digital playback but with analog as well. Unfortunately, there is little difference in the results. The 'live' guitar sounds very different than the reproduced. I have also done this test with tape....slightly( VERY slightly) more convincing, however, still not able to get the reproduction absolutely correct. Nothing, I repeat NOTHING, is able to reproduce the sound of a 'live' guitar so that it sounds like the 'real' event just played, IMHO. If someone thinks this is NOT correct, I would venture to say that that someone just hasn't heard a 'live' guitar played up close or perhaps isn't hearing too well, OTOH I suppose FMMV. ( Frank's mileage may vary)

So if we can't capture an instrument like a guitar that for arguments sake is essentially frequency response and dynamically limited, what does that say for other instruments?

 

[MylesBAstor]

Very profound questions, neither of which seemingly has been or will be answered by those on their own soap boxes who diligently advocating the superiority of one over the other, science be damned. Golden ears trump science.

No we haven't developed the measurements nor is it certain that these measurements will be of any usefullness given the wide disparity in inter-aural hearing abilities among humans.

 

[FrantzM]

No we haven't developed the measurements nor is it certain that these measurements will be of any usefullness given the wide disparity in inter-aural hearing abilities among humans.

?????

 

[Orb]

Taking this and combining it with the original point of that started this thread, what do we think the measurements show of a tube amp that doesn't sound like SS and one that does? Is there nothing revealing in there?

If looking at the few ideal products in both camps, then they measure very closely if also accepting current audible perception conclusions, such as the work presented in AES and other publications.
One example for tubes would be the ARC Annivisary preamp and the McIntosh Anniversary special edition 275 tube amp (from what I remember subtly different to the standard model and only sold with McIntosh matching preamp), they match solid state.
That said it would be easier differentiating between pre and power amps, where possibly the preamp is easier to find more products that match well in terms of traditional measurements, though the traits of a tube preamp is not as great as say the best of their power amps but this also requires well restricted use in terms of speaker sensitivity-impedance-etc.

Anyway on testing-subjective listening, the problem IMO is that if trying to correlate sound differences, then real music complex "tone" is required, as an instrument sound is identified by its attack (critically time domain) and amplitude behaviour of its fundamental/partials-harmonics, while the sustain and decay will also affect the listener subjectively but in different ways.
A sinewave will never help to really correlate the behaviour of the best amps.

And this is where I have to differ from Tim about further testing, it is about testing a very well recorded instrument and major chord as it is in output from the player, and also in output from the pre and power amps.
A sinewave will never replicate what music (even if it is just a major chord) does in both the frequency and time domain.

Interestingly from what I have read if an instrument is just played louder then to us it sounds the same as the note just played, but in fact the two waveforms are different even when loudness is accounted for and they are compared again.
This shows just how much more complex musical notes are to a sinewave.

Anyway such tests that I mention are meaningless in building-testing an amp, but would be useful in comparing two products that are similar and yet preferences can and do differ between the products.
Possibly would be intersting to compare cheap mass produced high negative feedback amps to those that measure as well with no global feedback using such test, cheap class AB to a very good Class A and where both measure the same.
Edit:
Just want to emphasise that such test needs to look at the fundamental-harmonics from a scientific view of what is sound, which shows fundamental-harmonics with amplitude against time and provides a detailed envelope/spectral decay of the sound.
A snapshot of the fundamental-harmonic in the usual fourier analysis is not ideal as it does not show us the complete time domain of the sound.

Cheers
Orb

 

[Phelonious Ponk]

Myles, I have done this test with not just digital playback but with analog as well. Unfortunately, there is little difference in the results. The 'live' guitar sounds very different than the reproduced. I have also done this test with tape....slightly( VERY slightly) more convincing, however, still not able to get the reproduction absolutely correct. Nothing, I repeat NOTHING, is able to reproduce the sound of a 'live' guitar so that it sounds like the 'real' event just played, IMHO. If someone thinks this is NOT correct, I would venture to say that that someone just hasn't heard a 'live' guitar played up close or perhaps isn't hearing too well, OTOH I suppose FMMV. ( Frank's mileage may vary)

Regardless of Frank's mileage, it's not going to happen, and it has almost nothing to do with analog/digital/tube/solid state. At one end you have a guitar and a microphone, at the other you have a speaker and a pair of ears. They are radically different in the way they receive, respond, radiate...give me awhile, I'll come up with a couple more Rs. If you managed a perfect signal path between those two points, and you can get fairly close, it wouldn't sound like a live guitar. It doesn't have the physical properties for that. This is why the two most important things, IMO, you can do between those two transducers is leave the signal from the first one unmolested and provide sufficient control over the second one. The guy changing power chords in a system with weak driver control? Man, that dude is in over his head.

Tim

 

[Phelonious Ponk]

And this is where I have to differ from Tim about further testing, it is about testing a very well recorded instrument and major chord as it is in output from the player, and also in output from the pre and power amps.
A sinewave will never replicate what music (even if it is just a major chord) does in both the frequency and time domain.

We don't really disagree, Orb. I don't believe everything can be tested. I certainly don't believe I can look at a graph of a sinewave measurement and predict what something is going to sound like. I just tire of people taking positions against all the existing measurements, inadequate as they may be, in languge that sounds like all the science is on their side. And frankly, personally? I tire of it because I know what they're expounding is an opinion, because I disagree, not by measurments, but by listening. I find the quiet, the low distortion, the higher dynamic range and, specifically, most importantly, the faster transient response of digital and solid state to sound more natural.

MHO. YMMV. But the data that does exist backs me up, and I've heard an awful lot of real music.

Possibly would be intersting to compare cheap mass produced high negative feedback amps to those that measure as well with no global feedback using such test, cheap class AB to a very good Class A and where both measure the same.

This has been done several times, going back to the infamous Stereophile "Do all amps sound alike" test in the 1980s and Hafler's challenge. Did these tests prove that there is no difference between cheap solid state and expensive tubes? No. But they demonstrated pretty clearly that the differences are hard to identify at best. There is really no point in doing it again. You would just end up with the same believers and disbelievers.

Tim

 

[Bjorn]

A quote from Dr. Geddes which I find interesting:

"In my mind its not tubes versus transitors, its crossover versus non. Virtually all transitor amps are AB and thats a recipe for disaster at low level zero crossings which is exactly where things are the most audible. "

http://www.avsforum.com/avs-vb/showthread.php?t=1183019

 

[DaveyF]

So if we can't capture an instrument like a guitar that for arguments sake is essentially frequency response and dynamically limited, what does that say for other instruments?

Good point

 

[Phelonious Ponk]

A quote from Dr. Geddes which I find interesting:

http://www.avsforum.com/avs-vb/showthread.php?t=1183019

Dr. Geddes believes that once you get past a pretty low, inexpensive quality threshold, the differences between audio electronics are insignificant. Dr. Geddes points to other scientist-types who agree. Dr. Geddes supposes that the dramatic audible differences between audio electronics are promoted by builders and their marketing men. Dr. Geddes personal amp is a Pioneer receiver. Dr. Geddes is not popular with the high-end crowd.

Tim

 

[Gregadd]

So Tim is Dr. Geddes a believer or a scientist? I humbly submit if it were true he would be the most popular man in audio. Not even Pioneer would suggest such a thing. Tim In proof of which I refer to the TAD line of equipment.

Furthermore if you have not heard a high end musical system, it's not because they don't exist.

 

[Phelonious Ponk]

So Tim is Dr. Geddes a believer or a scientist? I humbly submit if it were true he would be the most popular man in audio. Not even Pioneer would suggest such a thing. Tim In proof of which I refer to the TAD line of equipment.

Furthermore if you have not heard a high end musical system, it's not because they don't exist.

I've heard many high-end systems, and I suspect Dr. Geddes has as well. One would think this kind of discovery would be popular, but in the high-end community it is enough to make a pariah of a scientist. For my part I'm not in the camp that believes that there are no audible differences between well-designed amplifiers operating well within their limits. I'm in the camp that believes the differences between well-designed amplifiers (and yes, by the way, I would put most mainstream midfi receivers under this umbrella) operating within their limits are so small that they're not likely to be reliably recognized once sight is removed from the listening experience.

That's pretty small.

With that said, "within their limits" can define a pretty big gap in this context, and the limits of a basic Pioneer receiver and some high-end big iron are dramatic. But Dr. Geddes designs very efficient speakers with a pretty even, easy 8-ohm load. I doubt they are a challenge to his Pioneer, which would probably break sweat just looking at your MLs. Horses for courses, as they say, but not in some diaphanous synergy sense, just a proper match of amp and load. The easiest, and most effective way around it, of course, is engineered active systems.

Tim

 

[Orb]

Dr. Geddes believes that once you get past a pretty low, inexpensive quality threshold, the differences between audio electronics are insignificant. Dr. Geddes points to other scientist-types who agree. Dr. Geddes supposes that the dramatic audible differences between audio electronics are promoted by builders and their marketing men. Dr. Geddes personal amp is a Pioneer receiver. Dr. Geddes is not popular with the high-end crowd.

Tim

I doubt scientists will have consensus on this for quite awhile
As an interesting example Paul Miller, John Atkinson, Keith Howard, are all science degree educated in Chemistry (and for two of them actually worked in scientific research and have a masters or PHD), and yet they have a different view on this.
To me these are as much scientists as some others mentioned in audio, and possibly more with their backgrounds in both education and actually working in scientific research for awhile.
Not adding Kal as not 100% sure if he fits fully with the views of these three on this subject matter.
And I am sure there are other contributors in audio publications who have scientific backgrounds, just rare that these are ever mentioned.

Also thanks for the additional info in response to my earlier post, makes sense but we possibly have a different perspective on such testing, which I agree would have limited scope but I feel useful in the way I commented.
Cheers
Orb