Jump to content



Sign up for a free account to remove the pop-up ads

Signing up for an account is fast and free. As a member you can join in the conversation, enter contests and remove the pop-up ads that guests get. Click here to create your free account.

Photo
- - - - -

Interesting tube vs transistor article


This topic has been archived. This means that you cannot reply to this topic.
50 replies to this topic

#1 of 51 OFFLINE   Yogi

Yogi

    Screenwriter



  • 1,741 posts
  • Join Date: Jul 25 2002

Posted May 08 2004 - 03:25 AM

Found an interesting discussion about tube vs transistor.

http://www.theabsolu...._vs_trans.html

There were a couple of interesting points raised about accuracy and measurements. Please dont start flaming.
The truth is not out there but within you.

#2 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 08 2004 - 12:46 PM

oops

#3 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 08 2004 - 12:47 PM

OK, I won't flame, instead I'll comment on various parts of the link and toss in my requisite amount of sarcasm here and there. BTW, hi! I heard you moved.

HP says...And the essential mystery, the big question behind the question, is why do they still sound different? Because they do. That Pearson and TAS seek to obfuscate the matter and advance preposterous theories simply shows their unwillingness to understand measurements or go beyone simple one-point specifications. The behavior of amplifiers with high output impedance (frequency dependant mind you) interacting with varying speaker impedance curves to result in frequency shaping of the musical waveform was known decades ago. Bell Labs recognized the vagaries of this as well as the different characteristics of tubes between manufacturers and as they aged. They knew this would result in a lack of consistency and hence developed negative feedback. However a lack of consistency is consistenly defined by the high-end wonks as revealing. And then we have issues with poorly regulated power supplies. Spin, spin, spin.

Jon Valin states...They have more bloom; they have more light; they have more body. They do that thing I call “action,” which solid-state doesn’t. Solid-state is like a slide show.
Well look at the impedance curve of a speaker. Something happening in the midrange? Yes, solid state, properly implemented, is a slide show. It shows what's really there. If you want it colored, then by all means color it. Just don't portray it as real.

But on the back half, the steady-state tone and decay, tubes just eat solid-state alive.
Are we talking musical reproduction here or Hendrix or Boston with Marshall amps and other processors?

...solid-state is extremely detailed in that sense. But it’s not detailed in the other sense. It doesn’t have the low-level dynamic and harmonic information, or the air.

Ummm...since when is adding distortion in playback considered information? Did this come from the DoubleSpeak Dictionary in 1984? Shades of George Orwell!

Harley...Electrons flow through a solid piece of semiconductor material in a transistor—there’s a physical piece of matter. In a vacuum tube, the electrons travel through a vacuum. You have this boiling cloud of electrons above the cathode that get pulled towards the plate, and there’s nothing to impede them. [i]I don’t know that this explains the difference in what we hear.[i]

No Mr. Harley it doesn't explain anything so you're right about that. Even a broken clock is right twice a day.

Pearson...What Robert just said is absolutely logical. Electrons have to travel through matter, and they’re doing electronic jumping, whereas when you have a stream of electrons going through a vacuum, that is entirely continuous. It’s like a waterfall, as opposed to water going through a sieve.
Yes, and after they 'jump', they're back in matter again. There's a confusion here about the movement of electrons in matter (drift velocity...pretty slow actually) as opposed to wave propagation. I guess none of these chaps will be taking freshman physics.

Harley...I’ve measured dozens of tubed and solid-state amplifiers, and if you look at the harmonic distortion spectra of them, tubes produce primarily second and third harmonic components, but transistor distortion is more upper order, such as seventh and ninth harmonic. Some citations would be appropriate here. However consider that even if true, generally the absolute magnitude of transistorized distortion even in the higher orders is quite small and can well be masked by any musical content that occurs there. It's much much easier to hear though if you're dealing with pure test tones. The relative absence of even order distortion components in transitorized devices is invariably a consequence of applying feeback and such.
Here's the IM distortion of a Bryston amp...probably less costly than an equivalently powered tube.
Posted Image
Wow...IM distortion is down over 100 dB...Good luck hearing it.

Pearson...And, you know, when tubes do that, they give you the sense of having much more power. A 60-watt solid-state and a 60-watt tube amp never sound equivalent in terms of power. Poorly regulated power supplies coupled with speaker/amplifier interactions as previously mentioned have something to do with that. I wouldn't be bragging if I had a 60 wpc amp.

Jon: So you’re, you’re saying that tubes smear detail? That they have lower densities of information?

Paul: Why don’t we just say smooth? I mean, you can use a more neutral word like “smooth,” which is in fact what I think is going on.

Jon: Well, it’s an interesting argument, and I’ve actually never heard that one made before—

Harry: You may never hear it again.

Jon: [Chuckles] But I’m not so sure I believe it.

Harry: I don’t.


Well then let's say tubes alter the information and in certain instances augment the signals. It's wasn't a safe thing in Galileo's time to challenge the Church either. Pearson feels better believing in things like there are electrons that carry musical information and those that don't.

Jon: If you’re talking broadband, then, yes, I think transistors do have more information overall.

Harry: And I think you would say that is more accurate overall.

Jon: I would.

Harry: It’s just not more truthful.


Martha Stewart could've used Pearson as a defense attorney.

Jon: [Laughs] Well, that’s what this magazine is about, Paul, making judgments on the basis of experience.

And damn reality while we're at it. It's about preferences and keeping readers in the dark ages.

Pearson...Okay, take the Edge [solid-state] amp. What do I hear? I hear an incredible cleanness, a top-to-bottom harmonic accuracy that’ll beat any kind of tube unit I know. Do I hear in the midband the harmonic and dynamic subtleties of the best tubes? Never. I think they subtract harmonic information and I don’t think it’s a matter of my hearing excessive second harmonics.
No, it doesn't subtract information, it just doesn't add it. It also doesn't frequency countour the signal with respect to the speaker impedance curves. Neither will an ATI or a Yamaha. If you want some bloom in your solid state amp, just add a 2 ohm resistor to your speaker wires.

Paul: I would be actually much more interested in trying to figure out, for example, why, on some kind of hard-edge, scientific basis, amplifiers sound the way they do. It would be really interesting, to me at least, to try to throw some things up, to get some good, rigorous measurement, and see if there’s some technical basis to this.

Robert: I can tell you right now, Paul, that nothing in the conventional set of measurements in use today would give you a clue as to why the amplifier sounded like that.

But he knows. And so do you. Run the measurements on a live load and watch the output impedance of the tube units go from 1.5 ohms to 20 ohms or more as frequency is increased. Let's not tell the natives the reason this glass globe lights up is because it's attached to a battery. It's much more fun to be mystical.

Harley...It’s been my experience, after running the Stereophile test laboratory for eight years, that it does not. Well thank god he's out of there. Just 'cause your the Chancellor of Germany and got reelected doesn't mean you've got a clue as to how to pull your country out of an economic recession.

In general this was like tossing in Sylvia Browne, Edgar Cayce (back from the dead), Van Pragh, Edwards, The Psychic Hotline, and Uri Geller into a room moderated by that hard hitting investigative journalist, Larry King. In the end all we have is chicken bones on the floor, but them hot wings were damned tasty. How does the Colonel do it?

#4 of 51 OFFLINE   PaulDA

PaulDA

    Screenwriter



  • 2,583 posts
  • Join Date: Feb 09 2004
  • Real Name:Paul
  • LocationSt. Hubert, Quebec, Canada

Posted May 08 2004 - 01:12 PM

I bought that issue (was looking for some entertainment AND they review music that's not always reviewed elsewhere, which I find useful). I noticed the only dissenter in the ranks re:tubes, is also the guy who says Home Depot speaker wire is perfectly fine for the 20000$ and up systems (though his editor felt the need to put in a disclaimer, to avoid angering the advertisers, I suppose). Anyway, Chu, as usual, your rebuttals are quite illuminating. I'd even go so far as to suggest there's some effervescent quality to themPosted Image .
Never try to teach a pig to sing. It wastes time, and it annoys the pig.

#5 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 08 2004 - 02:53 PM

I think that was HD extension cord. I'll be he doesn't get invited to any parties Posted Image

The simple fact that the high-end can't measure differences or tend to do so with tape measures should by no means be construed that the scientific community as a whole is guilty of the same level of incompetence. The high-end is guilty of the most egregious lies and deceptions and claims. They've waved around how 'they' discovered Transient Intermodulation Distortion when the facts indicate that it was known a 1/4 of a century earlier. So were the cures. The same holds true for jitter. Also discovered, studied, and understood as far back as the 50's by the telecommunications industry.

BTW, I hope it's understood I'm not critiquing tubes here. Simply the drivel that passes for information and masquerades as knowledge.

#6 of 51 OFFLINE   Yogi

Yogi

    Screenwriter



  • 1,741 posts
  • Join Date: Jul 25 2002

Posted May 08 2004 - 02:53 PM

Chu, I have moved to CA few months ago. Like it so far. Only thing is I am living in a cardboard box (cos its sooo expensive down here), which has excellent acoustical properties BTW, due to its non reflecting surfaces. In any case it was good to hear from you.

Quote:
If you want it colored, then by all means color it.


You and I both know speakers color the sound (more than a few db) more than anything else in the reproduction chain. Do you think you are listening to the real performace, uncolored? I didn't think so either. So accuracy is a figment of your imaginationPosted Image

Quote:
The behavior of amplifiers with high output impedance (frequency dependant mind you) interacting with varying speaker impedance curves to result in frequency shaping of the musical waveform was known decades ago.

Quote:
Well look at the impedance curve of a speaker. Something happening in the midrange?

Quote:
Poorly regulated power supplies coupled with speaker/amplifier interactions as previously mentioned have something to do with that.

Quote:
It also doesn't frequency countour the signal with respect to the speaker impedance curves

The only point that you make (over and over in different words as one can see) in your whole thread is the speaker/amp interaction in tube amps and how it varies the FR over the audio spectrum. I suppose you are talking about SET amps which I agree sound different with every speaker and don't sound right (but sound seductive) with any speaker. Modern push-pull tube offerings from CJ, ARC etc don't modulate their impedance with most speakers. For example the latest premier 140 from CJ stays within 1db with real speaker loads over the whole audio band. So it is well within the range of accuracy of most speakers. A modern push-pull tube amp mated to an appropriate speaker load (like the Von Schweikert VR2 whose impedance doesn't drop below 7ohms) would rival the accuracy of any SS amp/speaker setup. Tube amps have to be mated to the right kinds of speakers to work their best but once that synergy is achieved they are much more musically satisfying and just as accurate as a SS/speaker setup. Enough said about accuracy. Here again is your point quoted ad nauseum:

Quote:
Run the measurements on a live load and watch the output impedance of the tube units go from 1.5 ohms to 20 ohms or more as frequency is increased

Definetly sounds like a SET amp. My CJ stays within 1db over the whole band, but then its a push-pull amp.

Quote:
And damn reality while we're at it. It's about preferences and keeping readers in the dark ages.


Exactly my point.Posted Image It definetly sounds like someone is trying to keep the people in the dark ages when it comes to modern day tube amps.

Quote:
Simply the drivel that passes for information and masquerades as knowledge.

Exactly, its this drivel (about variations in output impedances) that was true in the 60's and that passes along as information for so long that people start thinking its a fact.

Cheers,
The truth is not out there but within you.

#7 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 08 2004 - 03:08 PM

Not to say that CT is the cheapest place in the world to live either. Must've been one hell of an offer to go to Cali. How's the missus like it?

SETs are but one example. OTL's are another. I'd love to see the output impedance vs. frequency plots if you've got them Yogi.

#8 of 51 OFFLINE   Yogi

Yogi

    Screenwriter



  • 1,741 posts
  • Join Date: Jul 25 2002

Posted May 08 2004 - 03:22 PM

It was a good offer at a startup company with a ton of options. But again as you know with startups if it works then its great and if it doesn't I'll be looking for a job couple years down the road. The missus doesn't like it much as she had all her kitty parties in CT and now there are nonePosted Image The kids luv it though.

Actually you could check out the soundstage review (I'll post it if I find it) of the CJ Premier 140 where they have a measurement of the output impedance and from what I rememberits impedance didn't vary by more than 1db.

Yes, I agree with your points WRT SETs and OTLs.

Cheers,
The truth is not out there but within you.

#9 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 08 2004 - 03:30 PM

I'll cross my fingers for you Yogi. Economy is looking better and better. If Hillary gives me a stock tip, I'll pass it on. Best of luck Posted Image

#10 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 09 2004 - 12:15 AM

Here's the FR of the Conrad 140 (a nice amp and certainly IMHO, a better than your typical performer) into a simulated NHT load
Posted Image

Now here is that simulated impedance load.
Posted Image
The red trace is the impedance while the magenta is the phase. Note the correlation between the Conrad's FR and that of the simulated load.

This can be contrasted with a run of the mill solid state amp like the Anthem PVA2.
Posted Image
Soundstage doesn't even bother putting the simulated load of the graph because to see the effects, one would have to tremendously expand the scale.

Since the vast majority of crossovers occur in the midrange area, and since it is the midrange area that our hearing is most sensitive to variations in level, is it really any wonder that a tube amp can have a noticeable difference? It certainly appears so to Pearson and Harley and to magazines and publications cut from the same bolt of cloth as TAS.

BTW Yogi, how's the pizza in your neck of the woods?

#11 of 51 OFFLINE   Yogi

Yogi

    Screenwriter



  • 1,741 posts
  • Join Date: Jul 25 2002

Posted May 09 2004 - 03:41 AM

Chu how do you post pics in your thread. Also whats the y axis on the second plot? is it ohms?

Pizza sucks in Cali. But my fav food (Chinese) is heavenly in San Francisco. Chinese food in CT sucked big timePosted Image NY's China market was good though but not as good as the China market here.
The truth is not out there but within you.

#12 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 09 2004 - 04:53 AM

Yes it's ohms. To get those pics, which i got from Soundstages website, I simply right clicked on the images that I was interested in and got the image properties. I then copied that and used the IMG icon on top of the Reply area.

Seeing as you're in the Frisco area, you might as well look up Wyatt Earps tomb and pay old Wyatt a visit.

#13 of 51 OFFLINE   Lewis Besze

Lewis Besze

    Producer



  • 3,134 posts
  • Join Date: Jul 28 1999

Posted May 09 2004 - 05:46 AM

Quote:
Pizza sucks in Cali
Is this based on on your extended "vacation stay"? Posted Image

That article was more of a propaganda then a discussion,and certainly nothing new that we haven't heard.

#14 of 51 OFFLINE   David Judah

David Judah

    Screenwriter



  • 1,479 posts
  • Join Date: Feb 11 1999

Posted May 09 2004 - 07:39 AM

It just seemed like they were trying to put into words their impressions of the differences between tubes and SS, as imperfect as that may be(like trying to describe a dream to someone).

I don't get this whole notion of accuracy. We hear alot about it, and some expound the virtues of trying to acheive it, but considering how most music is made and recorded, I don't know how we could have much hope to reach that lofty Platonic ideal in the Forms.

So tubes and SS don't always produce the same results on a graph--who cares? You gotta go with what sounds best to you with the related equipment that you have.

Incidently, I don't own any tube gear, but I'm very curious about it. I've considered a tube CD player and a Behringer unit that is supposed to simulate a tubey sound, but I've had other priorities and haven't pulled the trigger yet.

IMO, there's got to be a little mysticism left for audio to keep things interesting.

DJ

Lecktor: Then how did you catch me?
Graham: You had disadvantages.
Lecktor: What disadvantages?
Graham: You're insane.

#15 of 51 OFFLINE   Yogi

Yogi

    Screenwriter



  • 1,741 posts
  • Join Date: Jul 25 2002

Posted May 09 2004 - 09:10 AM

So let me get this straight, Chu: you are saying that one can hear a +/- 1db perturbation in amps output with speakers having +/- 3db error over the audio band to determine that a ss/speaker system is more accurate than a tube/speaker system? I understand the reasons for differences in sound of both types of amps but when someone claims one system is more accurate than the other its confusing. What is accuracy? Maybe I am missing something.

Is it time for beer yet?Posted Image

AAH Lewis I didn't notice you are from Cali. Perhaps I have never been to a good Pizza restaurant in Ca. I'll keep searching.Posted Image
The truth is not out there but within you.

#16 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 09 2004 - 10:15 AM

In the amp you suggested, the CJ, between 2 & 4 kHz there is an approximately 1.8 dB swing. One could take a reasonally good stab at simulating that with an equalizer. Now if the equalizer has a defeat switch (pass-through) and you were to toggle that back and forth, most assuredly you would hear the difference.
I'll offer the following as my opinion. I think one of the more important things for a speaker to get right is a flat midrange. Let's look at the FR of a fairly well regarded, and not overly costly speaker: Ascend CBM-170.
Posted Image
The on-axis response is pretty good. In fact, the speaker itself is spectacularly flat from 100 Hz on up. It should do a quite credible job of reproducing the midrange accurately, something I prefer. Now I could certainly add some sort of tube based device and (let's assume for the moment the impedance curves happen to be identical to the simulated NHT load) and have the effect of boosting the 2kHz about 0.6 dB and dropping the 4kHz 1.2 dB. The effect would most definitely be audible. Significantly so? Depends on the individual. Could one get used to it? Well sure.

People who do reviews, and I'm not saying to necessarily trust them, often speak favorably with speakers that have a nice flat midrange. How do you think they'd speak if they saw a FR that mimicked that of some tube amp? However, alter the FR with tubes and out come the glowing adjectives and adverbs. Am I the only one who sees this as more than a bit two-faced?

Talking further along the ideas of preference and philosophy, I happen to not like the idea of people, a company, imposing their idea of equalization on me. I'd rather have my equipment flat and insert an equalizer into the signal path and if I want to alter things, I want to be in control of that. I also don't like the idea of tubes aging, having to second guess the bias, matching tubes, etc. I also don't like what I consider paying extra for these things I don't like and then not getting the power I desire. Call it my idea of bang for the buck. I like, if I replace an amp, that the new one sounds just the same as the old one.

So in summation, yes, you could hear that. Depending upon how someone spun what was happening, like those guys in the review article and depending upon one's susceptibility, you might like it or you might detest it. I just happen to like a little more control in this out-of-control world. That's just me. It doesn't have to be you.

#17 of 51 OFFLINE   David Judah

David Judah

    Screenwriter



  • 1,479 posts
  • Join Date: Feb 11 1999

Posted May 09 2004 - 10:54 AM

Quote:
I just happen to like a little more control in this out-of-control world

Then you need one of these.

DJ

Lecktor: Then how did you catch me?
Graham: You had disadvantages.
Lecktor: What disadvantages?
Graham: You're insane.

#18 of 51 OFFLINE   gregD

gregD

    Second Unit



  • 420 posts
  • Join Date: Jan 16 2003

Posted May 09 2004 - 11:33 AM

Pizza sucks in Cali. But my fav food (Chinese) is heavenly in San Francisco.

Yogi

Welcome to CA, where 'startup' is a dirty word... hope it works out for ya.

If you're in SF, try these for pizza:
Tradtional: North Beach Pizza... Haystack Pizza.
Not-so-traditional: Panhandle Pizza... Vicolo.

You'll need your strength to keep up these VT/SS threads.

Fun read... thanks.

#19 of 51 OFFLINE   Yogi

Yogi

    Screenwriter



  • 1,741 posts
  • Join Date: Jul 25 2002

Posted May 09 2004 - 02:50 PM

Chu you do still realize that the FR of that speaker also has almost a 4db variation even in the midband. If you put the CJ 140's response to that same scale it would be pretty darn flat. You are a technical person so tell me would you trust an instrument with an accuracy level of 4db to measure variations in something that has a variation of 2db? But I do see your point about the varying output impedance of a tube unit. But is that the only thing that gives the 'different sound'? Or is there something more going on there. I don't know so this is not a trick questionPosted Image

I think the 'tube sound' is a combination of varying output impedance, the even ordered distortions, the clipping properties of tubes etc etc. I dont think its as simple as adding an EQ to your chain.

BTW I am drinking a Heineken as I am typing this message.Posted Image Wanna crack one open?
The truth is not out there but within you.

#20 of 51 OFFLINE   Chu Gai

Chu Gai

    Lead Actor



  • 7,270 posts
  • Join Date: Jun 29 2001

Posted May 09 2004 - 03:06 PM

And I am having a Scotch for that warm glow on a breezy, chilly night.
I'm simply saying if we take a particular speaker, let's make it the NHT, power it with a SS amp, equalize it so that it matches as closely as possible the FR response of the CJ (boost here, cuts there...a 1/3 octave equalizer ought to do the trick) and then toggle the bypass, you will most definitely hear it. As to how big a thing you want to make of it, that's your call. If you can get a hold of an equalizer Yogi, give it a shot.
I don't know about the distortion thing of the CJ but I've got more than a sneaking suspicion that once you equalize it and then were able to rapidly switch between the CJ and the hypothetical SS amp (level matched), you'd be very hard pressed to tell one from the other.
I don't think I've ever driven my amp into clipping.


Back to Receivers/Separates/Amps



Forum Nav Content I Follow