[cupofwrathcom]

speaker designers are wrong because they made speakers that would be flat and ring less with such a stable amp. and I like how flat response is always equal to lifeless all along his page.

I'm not going to defend the article beyond the general conclusion that the main difference of tubes is the higher output impedance, same thing that Bob Carver said and did to match the sound years ago. The changes in frequency response are an effect of this. Transistor amps make sense on a lot of levels but not with certain drivers or speakers.

You can't increase the transients of certain drivers above a certain point, the box would be too small and you'd have no bass.
This article is heavy but very well written. The author designs a full range speaker that uses the higher output impedance of the tube amp.

http://www.diyaudio.com/forums/diyaudio-com-articles/158899-arpeggio-loudspeaker.html

maybe I'm missing something, but right now I don't see why it shouldn't be possible to simulate most tube amp behaviors digitally.

All i know is that it wouldn't change the physical q of the system. Maybe its good enough for some, this is probably what tube preamps do, but the damping is the same.

 

[pinnahertz]

I'm not going to defend the article beyond the general conclusion that the main difference of tubes is the higher output impedance, same thing that Bob Carver said and did to match the sound years ago. The changes in frequency response are an effect of this. Transistor amps make sense on a lot of levels but not with certain drivers or speakers.

Recall that Carver's goal was to get a 70dB null between the reference tube amp and his 1.0 amp, which ultimately resulted in identical amp sounds, though the Stereophile guys wouldn't permit ABX testing. The problem with getting a 70dB null between two different amps is that it's hypersensitive to inaudible differences like minute phase shifts and microscopic changes in amplitude vs frequency. Most of what he did was raise the output Z of his SS amp, but there was obviously a lot more to get that null. The 70dB null may not have been necessary, and the comparison tests were fully sighted and biased.

What does this have to do with tube amp modeling? Well, if you're trying to model a particular tube amp working into a particular speaker load, it's unlikely you'll find a plugin to do that. It makes no sense anyway unless you have the exact speaker to present that load. Moreover, that speaker is just the middle, actually kind of the beginning of a signal chain that includes the room. Modeling, therefore, would have to include the reference amp, reference speaker, reference room and reference LP. Today that is entirely possible, but rather impractical given the massive combinations of speakers amps and rooms available.

That's not what tube modelling tries to do. What it does is model the general characteristics of tube amps, even some specific ones, and they do that well enough.

This article is heavy but very well written. The author designs a full range speaker that uses the higher output impedance of the tube amp.

http://www.diyaudio.com/forums/diyaudio-com-articles/158899-arpeggio-loudspeaker.html

Yes, and he's a favorite author here too.

All i know is that it wouldn't change the physical q of the system. Maybe its good enough for some, this is probably what tube preamps do, but the damping is the same.

I'm not convinced that you couldn't model a highly damped system, and build a model that produced a simulated under damped result, and vice-versa. You would have to characterize both, and use a good SS amp. There would be issues, measurement would be tricky, but as long as you didn't exceed the physical capabilities of the speaker playing the model it could be done. The rough parallel would be phase-compensating a complex filter with an inverse all-pass network. Not quite the same thing, but it can be done in the analog domain, so it can be done better and with more precision in the digital domain.

I'm not sure why anyone would want to, though.

 

[cupofwrathcom]

but it can be done in the analog domain, so it can be done better and with more precision in the digital domain. I'm not sure why anyone would want to, though

I would consider building a speaker like the Arpeggio, but I would use series resistors to get the 6.5 ohm impedance instead of a tube amp. THen equalize it out with something like equalizer APO. Good tube amps are too expensive and high maintenance for my tastes.

 

[100VoltTube]

My 2¢: a badly designed tube amp can be accurately modeled (if you want that), but an amp that uses the advantages of tubes, such as lower grid vs. gate/base capacitance (which doesn't matter if the topology doesn't use that element as the input), triodes' linear performance without any negative feedback, and is well-designed overall cannot be modeled, as it would be higher performance in some areas than the SS amp that is emulating it. The 2nd order distortion will almost always be higher in tube amps, but higher order harmonic distortion levels will be higher in the SS amp due to the negative feedback than the well-designed tube amp being emulated.

 

[Whitigir]

There is no way to replicate tube sound exactly, the mechanism of how electron cloud is moving and reacting are too different from transistors to tubes valves, even though on a basic principle they are the same. One example is that Transistors can be fried and interrupted by EMP, while tubes valves are not, or barely affected at all. It is said that even after a nuclear fallout, any physical good looking tubes valves devices will still be functioning !

 

[Strangelove424]

There is no way to replicate tube sound exactly, the mechanism of how electron cloud is moving and reacting are too different from transistors to tubes valves, even though on a basic principle they are the same. One example is that Transistors can be fried and interrupted by EMP, while tubes valves are not, or barely affected at all. It is said that even after a nuclear fallout, any physical good looking tubes valves devices will still be functioning !

Tubes are immune from nuclear fallout because the materials are vacuum sealed and not exposed to the environment. I believe this was one of the reasons the Russians kept using tubes in military application long after the introduction of transistors. But I'd be far more concerned about knocking over my tube amp and breaking the tubes than I would be afraid of my SS amp not working in a nuclear armageddon. It's fascinating the way electrons are excited with filament and plate vs. across a semiconductor, but they're both pretty well understood phenomenon. Regardless, it's not electron movement across the amplifier that really matters, it's the end result of the signal that matters to a DSP. The audio signal or even the waveform can be examined, and any differences will be consistent. Those consistent difference can be described mathematically, and placed into a programming algorithm via a plugin. The basics of math and science say this is possible, maybe not well explored yet, but certainly possible.

 

[pinnahertz]

My 2¢: a badly designed tube amp can be accurately modeled (if you want that), but an amp that uses the advantages of tubes, such as lower grid vs. gate/base capacitance (which doesn't matter if the topology doesn't use that element as the input), triodes' linear performance without any negative feedback, and is well-designed overall cannot be modeled, as it would be higher performance in some areas than the SS amp that is emulating it.

If you modeled both and created an inverse of the SS, you'd probably get pretty close.

The 2nd order distortion will almost always be higher in tube amps,

Depends on the topology, bias, etc. I don't think that's an accurate statement generally.

but higher order harmonic distortion levels will be higher in the SS amp due to the negative feedback than the well-designed tube amp being emulated.

[/quote]That one I completely disagree with. I've tested quite a few SS amps, I don't see higher order harmonics unless it's driven into nonlinearity. And that's not operating an amp correctly. Tube amps have generally have a slower onset of nonlinearity, so distortion increases with level, which is hard to think of as good, but also hard to hear. SS amps generally remain very, very clean until they hit the wall, then things get nasty very quickly. But again, that's not running the amp properly.

 

[pinnahertz]

Tubes are immune from nuclear fallout because the materials are vacuum sealed and not exposed to the environment. I believe this was one of the reasons the Russians kept using tubes in military application long after the introduction of transistors.

It wasn't the radiation contained in fallout, SS devices are immune to that too. It's the EMP from a nuclear detonation that takes out SS junctions, but tubes survive just fine.

But I'd be far more concerned about knocking over my tube amp and breaking the tubes than I would be afraid of my SS amp not working in a nuclear armageddon. It's fascinating the way electrons are excited with filament and plate vs. across a semiconductor, but they're both pretty well understood phenomenon. Regardless, it's not electron movement across the amplifier that really matters, it's the end result of the signal that matters to a DSP. The audio signal or even the waveform can be examined, and any differences will be consistent. Those consistent difference can be described mathematically, and placed into a programming algorithm via a plugin. The basics of math and science say this is possible, maybe not well explored yet, but certainly possible.

You have to be careful here. Yes, anything could be modeled, but it's darn difficult to sample the device in the right point sometimes.

 

[castleofargh]

There is no way to replicate tube sound exactly, the mechanism of how electron cloud is moving and reacting are too different from transistors to tubes valves, even though on a basic principle they are the same. One example is that Transistors can be fried and interrupted by EMP, while tubes valves are not, or barely affected at all. It is said that even after a nuclear fallout, any physical good looking tubes valves devices will still be functioning !

cant say if it's a joke or the worst self justification ever in the history of people who like tube amps? electron cloud and nukes, because hey, why not?

 

[100VoltTube]

Depends on the topology, bias, etc. I don't think that's an accurate statement generally.

True. I probably shouldn't have generalized as much as I did.

I've tested quite a few SS amps, I don't see higher order harmonics unless it's driven into nonlinearity. And that's not operating an amp correctly. Tube amps have generally have a slower onset of nonlinearity, so distortion increases with level

Interesting... I'd love to see those measurements, if you still have them.

 

[pinnahertz]

Interesting... I'd love to see those measurements, if you still have them.

Nope, they stayed with the company I did them for, but it's not hard to find some. I looked around for a few minutes, and found these.

Here's a review of an Emotiva XPA, the distortion vs level graph is down the page a bit.

To compare, here's a review of an Audio Research VTM200, again, the level vs THD graph is down the page a bit. The two illustrate what I'm talking about, and are fairly typical. It's not had to find other examples...lots of them.

Here's one of a McIntosh MC275 (tubes). As usual, scroll down, noting the 8 ohm tap THD vs output graph.

Here's an NAD 218. Note that when you see THD+N plots dropping as level goes up, that's indicating the contribution of noise to the measurement. We could do THD only now, it'll basically show THD as a relatively flat line until the maximum is reached.

Those are typical of what I measured. I did a bunch of pro sound amps, and a few classic tube amps, including a McIntosh, a couple of Marantz (the mono block and the stereo, can't recall the models), and some more recent reincarnation of Dynaco tube amps. They all follow the basic profile of tube amps THD rising with level slowly, always much higher than SS, and SS amps THD staying quite low until the onset of clipping.

 

[bigshot]

If there's a nuclear holocaust, I'm covered because I have a wind up acoustic phonograph and bunker full of dance band 78s from the 20s! It'll be Paul Whiteman and cockroaches while all you guys lay in a pool of your own blood.

 

[pinnahertz]

If there's a nuclear holocaust, I'm covered because I have a wind up acoustic phonograph and bunker full of dance band 78s from the 20s! It'll be Paul Whiteman and cockroaches while all you guys lay in a pool of your own blood.

Dark. Very dark.

I worked for a radio station for a while that was outfitted with a fallout shelter, provisions for a month, water for a month, a diesel generator with enough fuel for a month, and a radio studio in the shelter. All paid for by the government. We tested everything once a month to make sure it worked. Even checked the batteries in the Geiger counters. And we all knew that if operating from there became a necessity we had our choice of living a month longer than everyone else (and broadcasting music to them while they died of radiation sickness), or just going outside to play and getting it over with.

I seriously don't think Paul Whiteman would help much, but one of the transmitters was a tube rig. Most of the other gear would have been blown up by the EMP, though, so I don't know what would have worked anyway. The generator and the transmitter, but probably very little audio gear.

 

[DamageInc77]

I know within the guitar amp world, some solid state amps with software suites have been able to perfectly mimic tube amplification sound and pass double blind tests. So why not with a headphone amp?

 

[old tech]

It puzzles me what and why anyone would want to emulate a tube amp, unless it is for production (ie achieving a type of distorted sound eg electric guitar amps) or if there is a certain playback preference for subjective euphony over accuracy.

A high fidelity amp should be transparent, ie amplify the signal without adding or detracting anything else to the signal. A well designed SS or tube amp are both capable to achieving this without any colour or perceptible distortion, providing it is operating within its output and headroom parameters. The issue is that a tube amp achieving this goal is going to be far more expensive, inefficient and high maintenance than an equivalent SS amp, so there is no real point. Therefore, most high end tube amps are designed to colour the sound (ie the tube sound or a "signature sound") as a point of difference and to justify the higher cost.

Carver proved in the 1980s that the sound of any amp can be copied, often at much lower cost than a high end amp, through nulling the output signal. Nothing magical here, just audiophile myths.