Head-Fi.org › Forums › Equipment Forums › Sound Science › How is "tube sound" even audible in modern headphone amps?
New Posts  All Forums:Forum Nav:

How is "tube sound" even audible in modern headphone amps?

post #1 of 366
Thread Starter 

I've been looking at Schiit Valhalla 2 specs and this is what puzzles me: how is it even possible to hear the "tube sound" with current crop of super-low noise/distortion/output impedance OTL headphone amps? Which is a great engineering feat by itself, but if the alleged "tube sound" is produced by certain harmonic distortions, how is it even audible with measured THD figures being so low?

 

I'm writing this in "Sound Science" as the similar question in the main forum produced a ton of over-emotional and less so informative feedback, if you know what I mean. Many people got simply upset. And it was kind of expected, but I honestly didn't want to troll... Subjectivist crowd is very sensitive.

 

Quote:
THD: < 0.04%, 20Hz-20KHz, at 1V RMS, high gain (worst case)
IMD: < 0.05%, CCIR at 1V RMS, high gain (worst case)
SNR: > 97db, unweighted, referenced to 1V RMS, low gain
Crosstalk: < -71dB, 20 Hz-20KHz
Output Impedance:  14 ohms (hi gain), 3.5 ohms (lo gain)

Edited by madwolfa - 6/2/14 at 12:06pm
post #2 of 366

No amp is the same as the next ... having said that, most tube amps don't have anywhere near those measurements and that's by design.

 

As for those who do, my guess is: see tube, hear tube.


Edited by eugenius - 6/2/14 at 12:41pm
post #3 of 366
post #4 of 366
Thread Starter 

I'm ready to believe that a big tube OTL amp, driving a pair of low impedance speakers, gives all kinds of audible distortions (pleasant or not). Not the modern headphone amp.

post #5 of 366

Well if it maintains similar specs into a real load with headphones it shouldn't have a sound of its own.  I have measured a DIY OTL unit that had similarly good specs into a test load, but with any real reactence involved it fell apart into a non-flat signal transfer with some significant levels of distortion.  My guess being it had inadequate current capability.  Would like to have investigated further, but not my equipment.

 

I don't know if this will be the case with the Schiit Vahalla or most OTLs in general.  I do know of one popular power amp for speakers which is OTL and suffers the same problem with all, but the most gentle speakers attached.  With most speakers it has an obvious sound of its own for obvious measured reasons. 

 

Perhaps others can say if such is the case with headphone OTL's in general. 

post #6 of 366
Quote:
Originally Posted by elmoe View Post
 

Here you are:

 

http://www.theaudioarchive.com/TAA_Resources_Tubes_versus_Solid_State.htm

 

A very interesting read.


I only skimmed through the IEEE article, but there didn't seem to be an answer to the OP's question there. Part history lesson, part review of the current market and a smattering of truisms.
​I wouldn't expect them to publish anything helpful on the subject in the future, either, considering this is the kind of stuff they deal with these days.

 

The AES article is from 1972. The OP's question was specifically concerned with modern tube amplifiers that have distortion levels below the known levels of perceptibility.


Edited by limpidglitch - 6/3/14 at 9:09am
post #7 of 366
Quote:
Originally Posted by limpidglitch View Post
 


I only skimmed through the IEEE article, but there didn't seem to be an answer to the OP's question there. Part history lesson, part review of the current marked and a smattering of truisms.
​I wouldn't expect them to publish anything helpful on the subject in the future, either, considering this is the kind of stuff they deal with these days.

 

The AES article is from 1972. The OP's question was specifically concerned with modern tube amplifiers that have distortion levels below the known levels of perceptibility.

 

So you came into this thread to do what exactly? Quote my post and start an argument not pertaining to the topic at hand? Without even bothering to actually read the article itself?

 

:rolleyes:

post #8 of 366

I skimmed the article. It was impossibly long and tedious.
Other than that I tried to steer the conversation toward the original topic, by pointing out that musings on 50 year old tube designs weren't really pertinent.

What esldude mentions about reactive loads seems to be much more on the money. The measurements might simply not be very realistic, by not taking into account the bother a real transducer might give a tube amplifier.

post #9 of 366

It's a pretty short article. Not tedious in the least. You tried to stir the conversation on topic by speaking about something off topic and providing no further insight, well done.

 

The only article that's pretty old is the one from the AES, not that tube designs have changed much since mind you. I'll stop feeding the troll now.

post #10 of 366
Quote:
Originally Posted by madwolfa View Post
 

I've been looking at Schiit Valhalla 2 specs and this is what puzzles me: how is it even possible to hear the "tube sound" with current crop of super-low noise/distortion/output impedance OTL headphone amps? Which is a great engineering feat by itself, but if the alleged "tube sound" is produced by certain harmonic distortions, how is it even audible with measured THD figures being so low?

 

I'm writing this in "Sound Science" as the similar question in the main forum produced a ton of over-emotional and less so informative feedback, if you know what I mean. Many people got simply upset. And it was kind of expected, but I honestly didn't want to troll... Subjectivist crowd is very sensitive.

 

 

That output impedance is high enough to potentially cause audible issues, especially on high gain. Other than that though, I agree - if it actually meets those specs, it should sound identical to a well-designed solid state amp when driving most headphones.

post #11 of 366
Quote:
Originally Posted by limpidglitch View Post
 

I skimmed the article. It was impossibly long and tedious.
Other than that I tried to steer the conversation toward the original topic, by pointing out that musings on 50 year old tube designs weren't really pertinent.

What esldude mentions about reactive loads seems to be much more on the money. The measurements might simply not be very realistic, by not taking into account the bother a real transducer might give a tube amplifier.


I was familiar with the links from elmoe.  I think it is a worthwhile point.  One that hasn't changed in principle.  Which is often the case with good, real useful information or science. 

 

Boiling it down, when actually compared, if clean enough both circuits sound clean and the same.  Where they differ is when pushed over into distortion.  Some trip up audibly in bad ways with smaller overloads than others.  Some tube fans, of which I was long one, end up with the top 10 db or so bouncing in and out of that gentle overload area.  They sound different than clean, and not bad.  Bounce SS the same amount and it creates bad sound quickly. 

 

It lends credence to the idea that clean enough sounds clean enough and the same regardless of circuit components and levels of feedback.  Have audible lack of fidelity and those other factors cause the audibility to differ between the two camps.  Moral of the story is keep circuits in their linear area of operation. 

post #12 of 366

Other than power levels or compatibility or impedances, clean is clean no matter how you get there unless a bad design has taken hold or one is subject to excessive imagination.

post #13 of 366
Thread Starter 
Quote:
The best tube amps can be functionally about as good as solid state amps. They wouldn't measure quite as good, but the difference would be mostly outside the range of hearing. That said, I think that most of the reason people like the sound of tube amps is because of the slightly steeper high end rolloff. A lot of speakers and headphones have goosed treble to increase "detail". The rolloff on tube amps corrects for that a bit. But an equalizer would be a much more precise and flexible way to address imbalances in transducers.

 

- bigshot -

post #14 of 366
Quote:
Originally Posted by madwolfa View Post
 

 

- bigshot -

It's better to buy headphones that sound right in the first place than trying to force funky phones to sound better.

post #15 of 366

In Put Power

DC24V
Input Sensitivity 100mV
Input Impedance 100KOhm
Out-Put Impedance 20~600 Ohm
Gain 30dB
Frequency response 10Hz-60KHz +/- 0.25dB
Signal/Noise Ratio >90dB
Dynamic range 84.6dBA(300 ohm) 89.8dBA(33 ohm)
THD 0.016%(300 ohm) 0.45%(33 ohm)I
MD + Noise: 0.045(300 ohm) 0.42(33ohm)
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › How is "tube sound" even audible in modern headphone amps?