Stereophile's Jim Austin disagrees w Atkinson; says tubes have something that can't be measured

[fpitas]

Well -- there is that.
But I'm tellin' you -- it' s je ne sais quoi.

Well, perhaps. But the transformer may indeed be part of the equation. Nelson Pass thinks so and made some First Watt amps with coupling transformers that tube aficionados like.

FIRST WATT

https://www.firstwatt.com/pdf/prod_sit3_man.pdf

 

[theREALdotnet]

Take power supply noise. What positive attribute does it have?

It conveys a sense of homeliness and nostalgia? Or perhaps power and ability?

This scene wouldn’t work without it:

 

[Rick Sykora]

Decades ago, John Atkinson‘s helped Stereophile gain credibility with audio aficionados. Kudos @John Atkinson!

Anymore, only go there for the reviews with measurements. Austin may be carving out a new niche for Stereophile but is clearly more about serving himself than helping others. Stereophile can only ride on its legacy for so long as Austin’s position does not offer any sustainable differentiation vs Audioholics, TAS, SoundStage or any number of emerging internet influencers that will prostitute themselves for ad revenues.

 

[DMill]

Decades ago, John Atkinson‘s helped Stereophile gain credibility with audio aficionados. Kudos @John Atkinson!

I appreciated Stereophile pre-internet. Some good writers and Johns measurements were as good as it gets. I won’t hate on Jim Austin too much. He‘s captain on a sinking ship. Tick Tock and the crocodile is the internet and objective evaluation of equipment. I suppose he’s doing all he can at this point.

 

[pablolie]

In the next audio generation, AI will read the artists' and mixing engineers' mind and adjust EQ to their intent? Somehow?

Unless the "intuitive" camp comes up with a remotely rational reason why measurements can't supposedly capture "the entire truth", they should admit to bias in preferring certain deviations from the original.

 

[GXAlan]

Of course the effects of a tube amp can be measured. And measured trivially.

"Tube effect" as a category cannot because it's a made-up term.

The other analogy is to say that “tube effect” = “circuit effect.” You can create special effects in many ways, and the same tube can be used in different circuits for different special effects… all of which can be measured using conventional lab instruments but not necessarily standard “dashboards.”

 

[pablolie]

The other analogy is to say that “tube effect” = “circuit effect.” You can create special effects in many ways, and the same tube can be used in different circuits for different special effects… all of which can be measured using conventional lab instruments but not necessarily standard “dashboards.”

The funny thing is that there is no doubt whatsoever you can recreate the exact tube harmonics and distortion with "solid state" stuff. But if you advertized it that way it would convince not one single tube believer.

 

[GXAlan]

The funny thing is that there is no doubt whatsoever you can recreate the exact tube harmonics and distortion with "solid state" stuff. But if you advertized it that way it would convince not one single tube believer.

Yes, but if you wanted to recreate everything you have to account for the frequency response at multiple loads and the HD at different levels, rectifier sag, transformer effects, etc.

In a way, it is like the digital simulation of a piano. That’s easier than the simulation of a tube amp, but it’s still not trivial. (Look at something like the Roland V-Piano versus the way that everyone just uses wavetable synthesis+recorded samples).

Everything that end up creating the piano sound is trivially measurable but the actual measurements you need to perform to develop the model are not.

 

[amirm]

There is nothing about having a "tube amp" category that means it can't be measured,

??? Vast majority of claims about tube amp sound are unproven and fall in the category of imagination. Tube glows red and people think it sounds warm. If we eliminate this, there is hardly anything left of this so called "tube amp sound" to measure. For this reason, folks need to do a blind test first, prove what they say they are hearing (soundstage, sounding like live concert, etc.) and then we can talk about how to measure it.

 

[pablolie]

Yes, but if you wanted to recreate everything you have to account for the frequency response at multiple loads and the HD at different levels, rectifier sag, transformer effects, etc.

In a way, it is like the digital simulation of a piano. That’s easier than the simulation of a tube amp, but it’s still not trivial. (Look at something like the Roland V-Piano versus the way that everyone just uses wavetable synthesis+recorded samples).

Everything that end up creating the piano sound is trivially measurable but the actual measurements you need to perform to develop the model are not.

it is not a "simulation" bro, it is a scientifically proven and mathematically accurate reproduction of the original.

Can a recording "engineer" screw up a recording? Sure we see it all the time.

Can an audio equipment engineer design something totally incompetent? Certainly, but given the available toolset it is inexcusable.

 

[Spocko]

??? Vast majority of claims about tube amp sound are unproven and falls in the category of imagination. Tube glows red and people think it sounds warm. If we eliminate this, there is hardly anything left of this so called "tube amp sound" to measure. For this reason, folks need to do a blind test first, prove what they say they are hearing (soundstage, sounding like live concert, etc.) and then we can talk about how to measure it.

LOL this is exactly how many hifi reviews read, just replace "warm" and "sterile" with colors: "That sounds sooo red" or "a bit too blue"

 

[GXAlan]

it is not a "simulation" bro, it is a scientifically proven and mathematically accurate reproduction of the original.

I think we got our wires crossed. Agree 100% with your comment. Was just pointing out that you would want to do a lot of measurements to capture “everything” that could be captured when quantifying the deviation from straight wire with gain. There is no truth to the idea that the tube is reproducing anything from the recording more faithfully.

??? Vast majority of claims about tube amp sound are unproven and falls in the category of imagination. Tube glows red and people think it sounds warm. If we eliminate this, there is hardly anything left of this so called "tube amp sound" to measure. For this reason, folks need to do a blind test first, prove what they say they are hearing (soundstage, sounding like live concert, etc.) and then we can talk about how to measure it.

+1.

And in the blind test I posted here, it was very hard for people to hear differences other than the frequency response roll off. No tube amp sound to be found. That was with a resistive load.

And in the non blind test, with speakers, I found treble sweetness to be real. Then I was able to measure a signal dependent treble boost, which I thought was extremely cool (as a predominantly Class D guy) but the tube fans came out and said that my tube amp under test wasn’t a great representative of the genre… In a way, to the tube magicians, it’s not cool to explain the magic with a “simple” frequency explanation. To the solid state fans, like me, it’s super cool to see a “simple” frequency response explanation that also isn’t the same as a fixed tone curve and is dependent on the musical signal.

 

[SIY]

The other analogy is to say that “tube effect” = “circuit effect.” You can create special effects in many ways, and the same tube can be used in different circuits for different special effects… all of which can be measured using conventional lab instruments but not necessarily standard “dashboards.”

Or if I want an effect, I can do it in solid state. There's nothing magic about the active device from a "sound" end. Assuming some device actually DOES have a distinguishable sound.

 

[DonH56]

Yes, but if you wanted to recreate everything you have to account for the frequency response at multiple loads and the HD at different levels, rectifier sag, transformer effects, etc.

Virtually all the load effect can be simply recreated by adding a resistor in series with the output of a SS amp to mimic the tube amp's higher output impedance. Rectifier sag drops the rails at high power, same as in a SS amp. Transformer effects tend to limited frequency response, easily emulated with EQ, and added distortion (plenty of distortion generators around). I think emulating a typical tube amp is easier than most people think. A good tube amp sounds like a good SS amp so there's nothing to mimic, with the possible exception of load impedance.

The hard part is the distortion series which is a function of feedback and thus loop gain. That would be a more involved discussion but largely irrelevant for music. The "soft saturation" argument is mostly a myth as hard clipping looks much the same tube or SS, but I suspect many more tube amps are clipped much more often due to their limited power.

In a way, it is like the digital simulation of a piano. That’s easier than the simulation of a tube amp, but it’s still not trivial. (Look at something like the Roland V-Piano versus the way that everyone just uses wavetable synthesis+recorded samples).

Everything that end up creating the piano sound is trivially measurable but the actual measurements you need to perform to develop the model are not.

I'd rather emulate a tube amp from a clean SS amp than try to model a piano. The piano is much more complex from a signal-processing point of view. Organs are easier than pianos, for that matter (my wife plays both).

IME/IMO - Don

 

[JSmith]

Answer is simple: they don't have critical listening abilities to hear what is plainly and provably there as unwanted aberrations.

Agree... yet you would think this would be an important ability to learn if commenting on such matters.

I'll just pop this here for those Stereophile readers that may come to this thread;

http://harmanhowtolisten.blogspot.com/


JSmith

 

[Blumlein 88]

I've not read all the replies so hopefully not redundant in this one.

It was the question of did a tube amp pass thru something accurate SS amps couldn't pass or was tube sound a coloration that caused me to do series amplifier testing. You connect an amp to a power resistor or synthetic speaker load, tap that with a couple resistors to reduced level to unity gain, and feed it to another amp powering your speakers.

If something about accurate SS amps failed to retrieve information from your source signal you'd expect a tube amp as the DUT to sound like wire as the following SS amp would fail to portray all information. If instead tube amps were a coloration, you would expect the following SS amp to accurately portray that and sound like the tube amp. Cutting to the end, tube amps are a coloration. They don't dig out information in the source signal, they add to it. It may be a coloration that is preferred, but there is no unknown, unmeasurable quality involved. Reversing positions an accurate SS amp feeding a tube amp is undistinguishable versus a piece of interconnect.

Jim Austin needs to do this instead of spreading rubbish information to keep uncontrolled listening supreme over measurements. Listening by itself in such a test would show him he is mistaken.

 

[Short38]

Long time reader of Stereophile. Always read backwards from John Atkinson’s concluding paragraph in the measurements section. Always read any subjective reviewer’s opinions of amplifiers weighing at least 100lbs. I place the highest weight on amplifiers weighing at least 100lbs AND requiring new dedicated circuits. I place additional weight on the opinion of subjective reviewers listening to passive loudspeakers weighing at least 150lbs per side containing custom speced ScanSpeak drivers. But mostly I listen to Amir’s points about measurement.

 

[beefkabob]

I just want to make it clear that the opinion expressed by JA1 here, though very well-supported, is not universally shared.
... there's another school of thought--embraced by certain other Stereophile writers--that believes that something less tangible is retained in some such amplifiers that is lost ... yet (and this is my opinion, as the magazine's editor), in a magazine committed to subjective experience--to listening--above all else, such opinions must not be dismissed out of hand.

Translation: Magic is real, folks. Santa is real. The Easter Bunny is real. Your feelings dictate reality. Science is for jerks.

 

[GXAlan]

Virtually all the load effect can be simply recreated by adding a resistor in series with the output of a SS amp to mimic the tube amp's higher output impedance. Rectifier sag drops the rails at high power, same as in a SS amp. Transformer effects tend to limited frequency response, easily emulated with EQ, and added distortion (plenty of distortion generators around). I think emulating a typical tube amp is easier than most people think.

LOL, I agree 99.9% of what you said. I think the “easier than most people think” is because what is easy for you is harder for ordinary people.

I think it’s easy to think that you can just download Equalizer APO and pick a preset (easy) while there are four variables to play with and a mix or hardware (resistors) and software (EQ)

Or if I want an effect, I can do it in solid state. There's nothing magic about the active device from a "sound" end. Assuming some device actually DOES have a distinguishable sound.

Agree. The mistake that even or odd harmonic distortion profile is attributable to solid state or tubes is still repeated. It’s all circuit. You can do voltage or current feedback for solid state, etc.

 

[beefkabob]

The funny thing is that there is no doubt whatsoever you can recreate the exact tube harmonics and distortion with "solid state" stuff. But if you advertized it that way it would convince not one single tube believer.

Musicians happily buy that stuff in software. Software is generally cheaper than an actual tube amp, and one piece of software can act like 1000 different tube amps.