Why people still use tube amps when there are plenty of tubes already used in the making of music

[zchopper]

Linearity is measurable as low distortion.
SS amps in general measure better than the best tube amps. Those tube amps are never single tube designs and have lots of feedback (local and overall) often including extra windings or on just one of the TX taps.

Its all in the implementation of the active devices in a complete physical circuit.

To be clear, I'm not a schill for the tube amp industry. I can't handle the cost and maintenance and cost of good tube amps. I just don't deny that there is something substantive to the tube sound. I don't think saying that tube lovers like listening to distortion machines is necessarily correct. I have done some online hearing test things before where they introduce a signal which would be equivalent to 1%, .5%, .1%, .05% and lower distortion. Could be my age, but I have a difficult time hearing the .1% signal if i'm not imagining it, and below that I hear no trace. I give more weight to the linearity and imd graphs posted here than thd.

 

[solderdude]

I don't think saying that tube lovers like listening to distortion machines is necessarily correct

It isn't correct. Tube amp owners sure don't hear nor think of it that way. In general they factually are listening to distorted sound. Often mostly non-linear distortion and a bit of linear distortion.
Not all SS is low in distortion either and you can build a SS amp with a similar distortion profile by using J-FETs for instance. Those SS amp listeners are equally guilty of listening to distortion machines.

On top of that there are plenty of recordings that went through purposely designed distortion machines anyway.

And its not only about distortion numbers (at a specific level and frequency) but the complete distortion profile of the entire device as well as the stimulus (input signals).

 

[Rhamnetin]

My $0.02 is that, when well engineered, they can be more linear. They produce less initial problems to fix.

Are you exclusively referring to the use of tubes in an amplifier's gain stage, and not the output stage?

I'm guessing the vast majority of manufacturers making tube amps don't care too much about linearity, since to date all the most linear amplifiers are solid state. Schiit seems like they might care, and the Freya+ doesn't measure terribly, but the Kara is still their most linear preamp. Not that I disagree with you, I just want to see what the best possible tube (pre)amp is capable of out of curiosity.

 

[Chr1]

It isn't correct. Tube amp owners sure don't hear nor think of it that way. In general they factually are listening to distorted sound. Often mostly non-linear distortion and a bit of linear distortion.
Not all SS is low in distortion either and you can build a SS amp with a similar distortion profile by using J-FETs for instance. Those SS amp listeners are equally guilty of listening to distortion machines.

On top of that there are plenty of recordings that went through purposely designed distortion machines anyway.

And its not only about distortion numbers (at a specific level and frequency) but the complete distortion profile of the entire device as well as the stimulus (input signals).

This!

Also, I don't think that there is such a thing as "the best tube amp". Pre or otherwise. As tube amps generally have different distortion profiles, it is pretty much subjective and down to what you like personally surely. Some like the SET sound more than ultra-linear. Others vice versa. Most likely the same with tube compressors, microphones and any other tube device. They are intrinsically not as high fidelity as SS, but as previously stated there are also class A SS amps designed with certain distortion profiles too...

 

[zchopper]

Are you exclusively referring to the use of tubes in an amplifier's gain stage, and not the output stage?

I'm guessing the vast majority of manufacturers making tube amps don't care too much about linearity, since to date all the most linear amplifiers are solid state. Schiit seems like they might care, and the Freya+ doesn't measure terribly, but the Kara is still their most linear preamp. Not that I disagree with you, I just want to see what the best possible tube (pre)amp is capable of out of curiosity.

Gain stage could be the output stage or not. At minimum the output stage provides current gain, but could be voltage as well. The answer is I don't know. I'm hypothesizing about why people still use tubes. I believe it's important that the amplifier be as linear as possible across the audible spectrum before feedback is applied and that will lead to less unintended consequences in real world situations and better sound. I think the multitude of graphs posted here triangulates what good engineering is and keeps manufactures honest, but as this website has grown it's influence, there is a definitive benefit to try to game the numbers and be at the top of charts and tube amps will never hit .000x or sinad numbers. But I can also appreciate what solderdude says and that all good/different sounding amplifiers are distortion machines.

 

[SIY]

From 0 volts to 10 volts or whatever your input signal varies between, the gain isn't constant.

That’s nonlinearity.

 

[Mordecai]

I work in sound actually (pro tools, film sound post production), and use tons of plug ins, so I'm very used to manipulating sound.

Many here have made your suggestion but it's highly unlikely an audiophile inclined towards tubes would adopt it. First, such VSTs/plug-ins are typically used for professional work integrated in to DAWs. Nobody that I know of has made a tube emulator plug in for easy integration in to a typical stereo system.

Also, even a plug in may not emulate precisely what someone likes in their particular tube amp.

Finally, your suggestion misses a big part of the appeal which, as mentioned earlier, is the physical tube amps themselves. Many find them quite pleasing both conceptually and aesthetically, as I do. They look beautiful when they are on glowing, and there's a satisfying "connection to audio history" in using such older style gear. That's part of the fun. It would be like saying "hey, you don't need a turntable or records, just get a VST to add surface noise and crackle to your digital music."

I second this. Tube amps have a visceral impact that’s hard to objectively quantify. Listening to a tube amp is an experience in itself. All the rich harmonics are there to be enjoyed. A tube amp’s non-linear response including saturation also aligns with how we perceive sound. I think people are confusing real world with the now heavily quantised, digital representation of it. We’re analog creatures, not digital robots.

The reason why people add ”tube effects” in production at one or many stages is that they want to capture that analog magic. However, tube sound cannot be perfectly emulated in digital format, at least not with current tech

 

[Adis]

There is a difference between production and reproduction.

 

[Mordecai]

Disregarding noise, or microphonics, or psrr, or any other possible negative which tubes can be faulted for. Just looking for a single amplifying element which by itself provides the most linear gain function, ie 1 volt in = X volts out across the entire audible range, it could be strongly argued (and has) that a direct heated triode comes out on top of that chart. Fixing non-linearities in a single gain device is the fundament engineering problem to be solved in every amplifier. You can try to match that device as close as possible and push-pull so some non-linearities cancel, run the device in class A where it is most linear, try to hold voltage steady, parallel multiple devices, lower the gain of the device, loop negative feedback, and a myriad of other engineering strategies to correct that initial issue. None of those fixes are free. They introduce new issues, additional complexities, and require increasingly adept engineers to deal with them.

My belief is there are few engineers with a passion for audio skilled enough to tackle the issues. We're left with simple tube amps with harmonics and blah blah distortion profiles, and clipping or whatever, where if you an suspend your disbelief and tune out the warts, the linearity of the device and lack of subpar engineering trying to correct those warts allows you to enjoy the purest fundamental amplified reproduction of the input signal.

I’m an electrical engineer. There’s nothing mystical about tube amps. But no sensible engineer would want to make a valve amp sound like a SS amp. Tube amps are left to themselves as the natural distortion from overdrives is what makes the sound rich and organic. If someone wants a clean sound, buy a cheap ass SS amp. Frankly, less capable engineers can easily make SS amps.

 

[antcollinet]

We’re analog creatures, not digital robots.

they want to capture that analog magic.

There is no such thing. You can take an analogue signal from your most "magical" analogue source, digitise it, convert it back to analogue - and if you are using competent ADC and DAC, the output from the DAC will be audibly indistinguishable to the un-digitised source - for everyone, even the most golden eared amongst us.

If you used the most sensitive measuring gear then you'll be able to detect tiny differences - but they will be way smaller than any human ear can detect, and way WAY smaller than the flaws from the original analogue source.

Sure - if you want to distort it with a tube amp - if that's what floats your boat : have at it. But that's not magic - it's just distortion.

 

[Mordecai]

I've seen a lot of justifications for tube amps over the years, from local DIY guy (who built me an amp in the 90s) to tube amp manufacturers and enthusiasts. And the proposition is that tubes have some inherent advantages in actual accuracy, where solid state is a series of "kludges" to corral issues of their own making (hence the allergy among some audiophiles to negative feedback and such). The idea that if you can do it right with a tube design, it can be better, a more accurate and "pure signal."

But I've never seen this ever shown in practice, in terms of measurements. Every tube amp review I can recall in Stereophile, for instance, may start out with the manufacturer's technical case for tubes, but the measurements show otherwise and they always measure higher distortion in practice vs any decent SS amp!
(Even if, in some cases, sonically they might be "good enough" to be indistinguishable from a decent SS amp. I think Roger Modjeski's Music Reference tube amps measured pretty well for a tube amp?)

So I just always presume that a tube amp will be at best (in terms of best practice design choices) indistinguishable from a decent SS amp or at worst audibly distorting. Until someone produces some impressive measurements otherwise.

Can the claimed benefits of low distortion, e.g. claimed for a single ended design, be realized if one uses just the right speaker perhaps, e.g. one that doesn't present the tube design with big challenges? High sensitivity, benign impedance?

Physics would always conclude that valve amps will distort at high gains. But it’s not a bad thing. It is what it is. Maybe it can match the clinical performance of a SS amp at low volumes and one can game the test numbers but where’s the fun in that? If that’s the use case, then buy a SS amp. The valve sound at high gain is just different. It’s musical, warm, complex, even sometimes unpredictable. It surely it does not appeal to ’purists’ who want their sound clean…nevermind that it’s practically musically less engaging than the distorted valve sound for as long as it looks nice on the graph.

 

[Mordecai]

There is no such thing. You can take an analogue signal from your most "magical" analogue source, digitise it, convert it back to analogue - and if you are using competent ADC and DAC, the output from the DAC will be audibly indistinguishable to the un-digitised source - for everyone, even the most golden eared amongst us.

If you used the most sensitive measuring gear then you'll be able to detect tiny differences - but they will be way smaller than any human ear can detect, and way WAY smaller than the flaws from the original analogue source.

Sure - if you want to distort it with a tube amp - if that's what floats your boat : have at it. But that's not magic - it's just distortion.

Yeah distortion. A musical distortion that people try to emulate digitally in various forms. Maybe these people are just…amateurs, certainly not audiophiles, right?

I have my best measuring gears. They’re attached to my head. And they’re quite good. I can hold a tune and I don’t need graphs to recognise an Ab played on various instruments.

 

[SIY]

Tube amps have a visceral impact that’s hard to objectively quantify.

In most cases, just the opposite. Output transformers limit deepest bass.

But no sensible engineer would want to make a valve amp sound like a SS amp.

Except that once you compensate for the higher output impedance (a tweak of EQ), they pretty much do. Unless you peek. The vast majority of nonsense about "tube sound" is from people who have never done a comparison with basic ears-only controls.

 

[SIY]

The valve sound at high gain is just different. It’s musical, warm, complex, even sometimes unpredictable. It surely it does not appeal to ’purists’ who want their sound clean…nevermind that it’s practically musically less engaging than the distorted valve sound for as long as it looks nice on the graph.

Try ears-only controls instead of reading (and unfortunately believing) stuff spewed from people with no experience in basic sensory evaluation, then get back to us.

Amps are not "musical," musicians are. All amps do is make small signals bigger. They are dumb pieces of electronics.

 

[Mordecai]

There is no such thing. You can take an analogue signal from your most "magical" analogue source, digitise it, convert it back to analogue - and if you are using competent ADC and DAC, the output from the DAC will be audibly indistinguishable to the un-digitised source - for everyone, even the most golden eared amongst us.

If you used the most sensitive measuring gear then you'll be able to detect tiny differences - but they will be way smaller than any human ear can detect, and way WAY smaller than the flaws from the original analogue source.

Sure - if you want to distort it with a tube amp - if that's what floats your boat : have at it. But that's not magic - it's just distortion.

Also, it’s always amusing when I hear lectures on DACs and ADCs and qualifers like ‘audibly indistinguishable’. I do hope that you really objectively know how these stuff work, the algorithms that are involved, the DSP circuitry required, the maths like FFTs, etc. I studied DSP in uni so I’m a bit more careful when it comes to the science stuff. I’d leave you be.

 

[antcollinet]

I have my best measuring gears. They’re attached to my head. And they’re quite good

I can assure you that (as measuring instruments) they are pitifully poor compared to an audio precision analyser

Not only can they not detect anywhere near the same levels of distortion and noise, they are subject to all sorts of random disturbances due to subconscious biases in the wetware between them - resulting in changes to your perception of the sound - even when the sound reaching them is unaltered.

 

[Mordecai]

Try ears-only controls instead of reading (and unfortunately believing) stuff spewed from people with no experience in basic sensory evaluation, then get back to us.

Amps are not "musical," musicians are. All amps do is make small signals bigger. They are dumb pieces of electronics.

Ouch. So listening to a tube amp isn’t a valid sensory evaluation? What is? Reading a graph on thd+n?
I play an electric guitar. I prefer tube amps. I hear the highly musical distortions at overdrive.
I also have a SS amp. i play the guitar the same solo, the same way…never had that valve sound. So the only thing I changed is the amp. 2 amps, 2 different sounds.
To say that all amps do is to amplify the signal is to deny reality. A crappy amp sounds like crap. A good SS amp sounds like an SS amp. A valve sound…sounds like how a valve amp should sound.
Somehow, it seems that joining this forum for some creates a reality distortion effect...like I have to deny what I’m hearing because it’s not supposed to be that way according to your graphs.

 

[Mordecai]

I can assure you that (as measuring instruments) they are pitifully poor compared to an audio precision analyser

Not only can they not detect anywhere near the same levels of distortion and noise, they are subject to all sorts of random disturbances due to subconscious biases in the wetware between them - resulting in changes to your perception of the sound - even when the sound reaching them is unaltered.

No, not really. They’re the gear that matter most. I mean, it’s not the precision analyser that’s supposed to enjoy the sound. It’s me.
Again, you are missing the point. Tube amps distort, they introduce harmonics. I get that. But I don’t give a rat’s ass because that’s a very pleasing sound. Of course, every now and then, the music breaks and it’s sometimes awful. But that’s part of the listening experience on a valve amp. It’s visceral, imperfect, and organic.

 

[antcollinet]

I play an electric guitar. I prefer tube amps. I hear the highly musical distortions at overdrive.

There is a huge difference between a guitar amp - used as part of the creative process, and specifically designed so it can be set up to produce massive amounts of audible distortion...

and an amp designed for audio reproduction - which is usually not.

 

[SIY]

So listening to a tube amp isn’t a valid sensory evaluation?

It is with basic controls. Ears only. No peeking. Matched levels.

No controls, no validity to an evaluation.