Tube Vs. Solid State (1 Viewer)

[mike_decock]

It is both astounding (to me, anyway) and wonderful that good performance can be had at such a reasonable price.

I've been tinkering in this hobby for 11 years and I didn't really figure this out until the last 2 years or so. I'm a happier audiophile than ever before. No doubt I will upgrade from my current setup, but I have a much better idea of what direction to take and where I will do my research (talking to people like Saurav on forums like HTForum and Audio Asylum).

-Mike...

 

[Saurav]

I'm a happier audiophile than ever before.

To me, that's the most 'accurate' gauge/measurement of whether your stereo is doing things right or not.

 

[Rich H]

My 2 cents about the tube vs. SS thing:

First, I have used many different SS amps and Tube amps in my system (for instance, I still use a Bryston 4BST sometimes and an older Harmon Kardon amp too). I tend to stick with tubes in my system (Conrad Johnson monoblocks, currently).

People (usually SS amp people) often say: SS for accuracy, tubes for warmth or "musicality." If by "accurate" one is talking strictly about reproduction of the electrical signal then, you can make a case for SS generally being more accurate (historically, anyway).

But one can also use "accurate" to mean and "accurate reproduction of the experience of live music." Framed another way: which one more easily fools my mind into thinking "that sounds like a real person singing?"

For me, in general, tubes do it better than SS amps. Most of us are most often exposed to sound as reproduced via SS amplification. I find one of the biggest differences between "real sound - e.g acoustic instruments and especially voices," and reproduced sound is that the reproduction sounds "mechanical" or "electronic." Through most sounds systems, voices take on a harder, spikier, less human texture (even the audiophile minimalist recordings).
Blame it on whatever we want: microphone coloration, recording location, mixing/mastering, speakers/room - whatever. The end result, whether the SS amp is accurately delivering the signal or not, is to my and many other people's ears, disappointingly stripped of it's organic sound. (I especially find the upper mids to sound thin, pinched, hard).

Organic, acoustic sound sources have a natural warmth to my ears that goes completely missing from most sound systems.
I find that music through a good tube-based system often sounds closer to what I hear from acoustic sound sources.
It sounds more appropriately full, especially in the mid/upper mids. Voices sound warmer, softer, fuller, fleshier - like real human beings do. Wood sounds more appropriately "woody." That kind of thing.

When I listen to music through my Bryston amp, there are many wonderful things I hear. But I virtually never feel that I could be hearing a real human being. Through my Conrad Johnson tube amps, it's an easy, regular occurrence.

I'm not making a single magical technical claim for tubes - if it's all distortion so be it. I can only say that, if tubes aren't producing for me an electrical signal that is more accurate, they are at least producing a sonic experience that is more accurate (or at least more analogous) to hearing real sounds.

Rich H.

 

[Lee Scoggins]

I'm not making a single magical technical claim for tubes - if it's all distortion so be it. I can only say that, if tubes aren't producing for me an electrical signal that is more accurate, they are at least producing a sonic experience that is more accurate (or at least more analogous) to hearing real sounds.

Amen.

And another possibility is that tubes may be more accurate on other unknown sonic measures while not having the lowest distortion numbers.

 

[Saurav]

And another possibility is that tubes may be more accurate on other unknown sonic measures while not having the lowest distortion numbers.

Good point. There are a few things wrong with using THD as a primary measure of accuracy. One, it's a static measurement, so it tells you nothing about how well the amp responds to instantaneous changes in frequency and power. Real music has much more of quick changes than long steady single-frequency tones.

Next, THD clumps all kinds of harmonic distortion together into one number. It's a well know fact that human hearing isn't linear when it comes to harmonic distortion - we will detect fairly low levels of 3rd order distortion, and find it unpleasant, but it's harder for us to detect 2nd order distortion, tests have shown that we don't hear fairly high levels of 2nd order harmonic distortion.

So, when someone prefers the sound of a tube amp which "severely colors" the sound because of it's 2nd harmonic distortion, most people conclude that the listener just likes that coloration, he doesn't want "true accuracy". However, it's possible that what he actually likes is the absence of the odd order distortion present in SS gear, because even though the level of that distortion is lower, maybe we humans can hear it more easily and it annoys us more easily.

At the end of the day, we buy our gear to listen to it, not to sit and measure it. So, to me, it makes sense to design gear that factors in the non-linearities in our hearing. For instance, (IMO) good equipment gives more importance to the midrange instead of the last bass octave. This is because we're more sensitive to midrange frequencies, so it makes sense to spend design dollars on getting that right, rather than sacrificing that in order to get response down to a frequency where we don't hear as well. Following the same logic, if we're more sensitive to one kind of distortion than another, doesn't it seem logical to design in a manner that reduces the more annoying distortion components, even if that means increasing the less annoying components as a compromise?

 

[John Kotches]

Saurav,

2nd order distortion is at the octave, Third is at an octave and a fifth.

The 2nd order is consonant, the 3rd order is dissonant.

An ideal solution has no compromise.

The catch is that 2nd (and 4th) order distortions makes the depiction "richer than reality", if that description makes sense. This regardless of whether the distortion is added to the fundamental or any of the naturally occurring harmonics produced by the instrument.

Regards,

 

[mike_decock]

The catch is that 2nd (and 4th) order distortions makes the depiction "richer than reality", if that description makes sense.

I understand this, it makes sense. It may not be an ideal solution but I like it more than making it "uglier than reality". My ears are more likely to believe "richer" than "uglier" and deceiving our ears is pretty much the goal, I think.

-Mike...

 

[Rich H]

John Kotches wrote:

.."The catch is that 2nd (and 4th) order distortions makes the depiction "richer than reality", if that description makes sense."

I hear you John. I'd retort that, since I find most stereo systems produce mid range/upper frequencies that sound "thinner and sharper than reality," tube amps can "re-balance" the sound somewhat. If my tubes are applying 2nd order distortion to thicken the sound where it is often too thin, so be it.

Although with my CJ amps, being a push-pull design, I don't really think it's 2nd order distortion. Probably some juicy transformer coloration or something of that nature.
I wish it were just an EQ thing, so that I could us my Z-Systems digital EQ to make my Bryston sound like my CJ tube amps. But there's more going on than just slight deviations in frequency response between my SS and Tube amps.


BTW, I'm a fan of Dunlavy designs too With good recordings, I find the Dunlavy speakers are one of the few brands that sound "even" - realistically filled out - throughout the mids. I've no doubt this is because of Dunlavy's adherence to flat frequency response, in contrast to the contoured mid range response found in many other high end speakers.

Rich H.

 

[Saurav]

All occurrences of the words 'harmonics' or 'harmonic' in my post referred to distortion added by the recording/reproduction equipment, not to the natural harmonics present in the instrument.

Later along in this CD, you'll be able to hear how much of each of various kinds of distortion are audible. It's not giving any secrets away to reveal that second-harmonic distortion—ie, the distorting component is adding a tone one octave above every note of the music—is inaudible even in large quantities. Fig.3 shows you why: a real instrument like the bass guitar already has large amounts of second harmonic present in its spectrum; adding a little more can hardly be expected to change the instrument's basic tonal quality. But because adding even small amounts of high-order harmonics changes the ratio of harmonics—hence the timbre of the instrument—by a relatively large amount, they will be more audible.

The table after this paragraph shows the distortion spectrum of an open E string played on a Fender bass guitar. With the fundamental frequency (41.2Hz) treated as 100%, the guitar already produces 390% of 2nd order distortion, 50% of 3rd order, 40% of 4th order, and so on.
Now, if you take all these harmonics and add 2nd order components to them, the results will still be even order withr espect to the original fundamental and the original harmonics. Add odd order to these harmonics, and your output spectrum will be all over the map.

 

[John Kotches]

Saurav,

Well this explanation is crappy:

The table after this paragraph shows the distortion spectrum of an open E string played on a Fender bass guitar. With the fundamental frequency (41.2Hz) treated as 100%, the guitar already produces 390% of 2nd order distortion, 50% of 3rd order, 40% of 4th order, and so on.

The author (or your translation) has confused naturally occurring overtones with distortion.

Regards,

 

[Saurav]

Well this explanation is crappy ... The author (or your translation) has confused naturally occurring overtones with distortion.

Nice of you to point out the one use of the word 'distortion' in my post that was incorrect, and ignore the rest of it
It was my error, not the author's, and thank you for pointing it out. I should have said "390% of the 2nd harmonic", not "390% of 2nd order distortion".
Here's what I was trying to translate, which is what I believe the author is trying to say - a Fender bass guitar already has 390% naturally occuring 2nd harmonic over the fundamental. So, if an amp adds another 3% or so 2nd harmonic distortion, it will be virtually inaudible over that 390% that is naturally occuring. Compare that to the third harmonic, where a much smaller percentage of distortion will be relatively more audible, since the naturally occuring overtone is only 50% in strength, which is almost 1/8th the power of the 2nd harmonic. Then throw in the fact that our ears will detect 3rd order distortion much more easily than 2nd order distortion (because, as you said, it is dissonant with the fundamental), and the perceived difference to our ears becomes even greater.
Hope that's clearer.
Anyway, that's a tangential example. Other musical instruments will have different harmonic structures, not all instruments have a 2nd harmonic that's 4 times stronger than the fundamental note.
My basic question still remains though - why would anyone pick a device that produced distortion that was (a) known to be more easily detectable by human ears, and (b) known to be more annoying to humans?
Take two hypothetical amps, both with 10% THD. One produces 10% of only 2nd order distortion and negligible amounts of other orders, the other produces 10% of only 3rd order distortion and negligible amounts of other orders. This is an extreme example, of course, but it serves to illustrate the point, and these 'amps' can easily be simulated using circuits or software.
On paper, both amps are equal, and have the same amount of THD. To our ears, the second amp is much worse than the first, because we find even 0.7% 3rd order distortion more annoying than 10% 2nd order distortion (as stated in one of the articles linked). Now let's say the second amp's THD were reduced to 9%, or even 5%. On paper it is now the better performer of the two, but to our ears it still sounds worse.
Which one would you pick? I know which one I'd pick, the one which lets a major chord still sound like a major chord, and not a weird maj-7th-9th-something (which is what I'm guessing will happen when you add 3rd harmonics to each of the 1st, 3rd and 5th that make up a major chord)
Replace those numbers with more realistic THD distributions of real amps, and the argument doesn't change, IMO.
And this isn't restricted to tubes vs. SS - design topologies, amount of feedback used, all of these affect the distortion spectrum of the amp, whether it's tube or SS. Even among SS amps, some have higher proportions of odd order harmonics, even though they show better THD numbers on paper.
Anyway, we've beaten round this bush enough, I believe I've explained my point adequately. And if after all this I still haven't, I should just shut up anyway

 

[Lee Scoggins]

Here's what I was trying to translate, which is what I believe the author is trying to say - a Fender bass guitar already has 390% naturally occuring 2nd harmonic over the fundamental. So, if an amp adds another 3% or so 2nd harmonic distortion, it will be virtually inaudible over that 390% that is naturally occuring.

Very interesting posts Saurav. If I understand this properly then we should be a lot less concerned about what Kotches said about even order harmonics on tube amps and more concerned about odd order harmonics on solid state. Is this correct?
I just feel that guitar chords and other acoustic music always seems to sound better on tube equipment both on the tube mic preamps we use in the studio and the tube powered playback system. Maybe this is why David Chesky, a trained musician, is always trying the best tube gear - his latest rave is Tenor Audio - and from what I've seen has a very hard time living with solid state. And think of the advantages of solid state in terms of lugging around to recording locations...
And it reinforces my view that THD is an almost meaningless number in and of itself.

 

[Yogi]

How about instruments that have odd overtomes naturally occuring? Would those instruments sound more natural with a SS amp? Or are there none that have odd order overtones?

 

[mike_decock]

How about instruments that have odd overtomes naturally occuring? Would those instruments sound more natural with a SS amp?

Most instruments have odd overtones as well, but in much, much lower amounts than even overtones.

However, I don't think that there are any instruments without even overtones, and they are usually in large amounts.

-Mike...

 

[John Kotches]

Saurav,

With respect to pointing out the flaw... It was a gross error and needed to be corrected lest people thought that an instrument could produce THD. Some instruments have effects pedals, but those are intentional distortions.

Then you followed it up with this:

My basic question still remains though - why would anyone pick a device that produced distortion that was (a) known to be more easily detectable by human ears, and (b) known to be more annoying to humans?

Take two hypothetical amps, both with 10% THD. One produces 10% of only 2nd order distortion and negligible amounts of other orders, the other produces 10% of only 3rd order distortion and negligible amounts of other orders. This is an extreme example, of course, but it serves to illustrate the point, and these 'amps' can easily be simulated using circuits or software.

Well sure, you've skewed the test in favor of the tube amp here. Why won't it sound better?

See above to look at more realistic results.

Regards,

 

[Larry B]

Saurav and John:
You guys can argue 'till the cows come home about which should be better, when we all know that each of us is going to listen to that which we like best.
But don't stop arguing on my account. :wink:
Larry

 

[Larry B]

Lee:

Maybe this is why David Chesky, a trained musician, is always trying the best tube gear - his latest rave is Tenor Audio -

I've been hearing a lot of great things about OTL amps lately and from what I gather, Tenor is perhaps the best of all of them. I was tempted to buy an OTL amp (Atma-Sphere, to be exact) but I was hesitant due to their not working well in all systems (impedance matching problems, etc.).

Larry

 

[John Kotches]

Guys,
As I pointed out above, Clarinet (which is an acoustic instrument last I checked) that produces odd order overtones. The "octave" key creates a first overtone that is the 3rd order (Octave and a 5th).
Saxophones have a two piece octave key, which is managed mechanically (thanks to Adolph Sax!) -- with the lower key @ the halfway point on the body, which covers fingered Ab down to fingered D and the upper key on the neck which covers fingered A through C# plus the palm keys. Without this concession Saxophone would sound like a loud, metallic clarinet. Hey wait a minute, that's what Kenny G sounds like
Flute has no octave key, overtones are @ the octave boundary and the different octaves are handled by embochure and air velocity. I don't recall the overtone structure for flute, but based on it being a "pipe with holes" suspect it's predominately even ordered.
Sorry, I don't have any insights on double reeds (Oboe, English Horn and Bassoon) -- I never could make a halfway decent sound on those instruments.
Brass -- a large contributer to the sound is the bore shape. Conical bore instruments (Cornet, Euphonium/Baritone, Tuba) have a warmer sound, ie richer in even order harmonics than "straight bore" instruments (Trumpet, French Horn, Trombone). Individuals can work on the sound of these instruments to alter it -- but here we're talking about the characteristic sound. Mutes also change the harmonic signature of trumpet and trombone as does the traditional "hand in bell" technique for French Horn.
Anybody that played brass instruments should chime in and help here -- I think that they normally start playing at the 2nd harmonic of the pedal tones.
Regards,

 

[mike_decock]

Wow,

Thanks for the lesson on brass and woodwinds, John! I'm a guitar player so I'm always thinking in terms of a string vibrating. Heck, with the right note under flourescent lights, you can actually see the harmonic vibrations of a plucked string. Kinda cool.

-Mike...

 

[John Kotches]

Mike,

I know a lot more about woodwinds than I do brass instruments, which is why I asked for some help.

Regards,