A good article about the human ear and tubes (1 Viewer)

[Yogi]

Thanks Scott for the suggestion. I have been thinking along the same lines of parallel SET with 300B output tubes model from Antique Sound Labs (model 1010 I think). I think I can get it used for around $1200. The Cary 300SEI would be more than twice the price (used) for me.

Ever wondered why SET amps are more expensive than their PP counterparts even though their designs are much simpler employing lesser number of components in the path?

 

[Steve_AS]

Has anyone really done any statistical hypothesis testing? Established a confidence interval? or a significance level for these kinds of tests?

Yes, Yogi, they have, in the audio hobbyist world going at least as far back as Greenhill's cable comaprison in Stereo Review in the mid 80's, Tom Nousaine's blind test reports over the years, a Stereophile blind test reprinted at their site (their statistics weren't very good, as it turned out, though) and very recently an ABX comparison of amps conducted by Tag McLaren Audio, to name some at random:

http://www.tagmclaren.com/great/sounds.asp

THen of course there's the articles that use DBT in the AES Journal and in professional journals of psychoacoustics, where blind test protocols are *standard* -- it'd be difficult if not impossible to get research involving audible comparison into such a journal *without* proper controls in place during hte experiments; the articles wouldn't make it past teh peer review process. It's been that way for decades, too, because perceptual bias has been known about for *that* long. Audiophiles sometimes pooh-pooh academic studies of hearing because they aren't tests of consumer audio components....which audiophiles seem to believe are going to behave 'differently' than lab-engineered components.

And then there's the fact that some compenent makers such as Harman-Kardon and Revel use DBTs in their products development; I'm pretty sure that they use standard statistical procedures to evaluate the results as well (see Sean Olive's recent work on speaker prefeernces, listed in the current issue of the AES journal)

If you confine your audio reading mainly to The Absolute Sound and similar journals, it's not surprising you don't see much about controlled comparisons, since the results of these would often make the reviewers in those journals look foolish indeed, and would make some of the advertisers rather extremely unhappy as well.

Side note to Chu: keep up the good fight! Your posts are always a pleasure to read.

 

[Steve_AS]

Fair enough from your point of view. However, there can be other arguments as convincing as yours. I, for example, think that the job for the playback gear is to recreate the sound of real instruments and human voices. That is lost once the recording process happens, so, replicating exactly what is in the recording is an absurd goal. What we want is to hear what was recorded, not the recording.

If it's 'lost' during the recording process, yet 'gotten back' during tube playback, why don't recording engineers 'get it back' using tube gear in the recording process, for the sake of us poor SS users? (Some actually do this, of course...it would seem to me that adding *more* tubes at the end of the line might be overkill in that case, though.)

I conclude from the fact that this is NOT the standard in audio production, that there cannot really be a consensus that so much is 'lost', or that tubes are the way to get it back, if it is. So to assert it as fact would appear to be premature.

 

[Yogi]

Steve, I dont think any one is claiming here that what is lost in the recording process is somehow regained in the playback system with tubes. All tubes do IMHO is make the music more involving and emotional. The human voices are more palpable (as opposed to dry) and immediate. They might be further distorting it or whatever and might not sound hi-fi at all but they sound musical.

I couldn't find the ABX test in the link you provided but will look more closely later. Perhaps I am missing something. Also statistical DBT for speakers would be pretty easy to establish which is what HK and Revel do, I suppose.

Cheers,

 

[Scott_N]

Yogi

I'm no expert but I bet some of the price difference is in the power supply and transfomers. Not to mention the fact a pair of WE300B's will set you back $700 a pair. I saw a 300SEI on ebay for $1700 a few days ago. You might also consider Opera/Consonance amps. Have you also thought about selling the MV55 and getting a push-pull that can switch between triode and ultralinear? Triode while not quite the same as SET is SET-like in that in stays in class A longer than ultralinear. The new R version of the Cary Rocket 88 has 30wpc triode and 60wpc in ultralinear and you can switch between them anytime.

 

[RichardHOS]

DBT of speakers is rather difficult compared to a DB comparison of sources, cables, preamplifiers, and even amplifiers. Not the least of problems is the physical size of many loudspeakers, and the inconvenience of having to physically move them to place each pair in the same position, and without the other speakers still being nearby producing unpredictable diffraction effects. Only slightly less irratating is the level matching for speakers.

Was it Harmon that built the quick-swap speaker apparatus to use for this purpose? Probably not a small or inexpensive piece of machinery.

 

[Michael R Price]

Is it just me or do we miss the point that solid state gear generates distortion too? It's not like any $500 100-watt stereo amp is perfect.

 

[Steve_AS]

Yogi, here's exactly what Manuel wrote, in the context of SS vs. tube amps:

". What we want is to hear what was recorded, not the recording.
And if any device (leaving aside any "accuracy readings") can do that then I will happily prefer it over other gear."

Was I wrong to infer that he feels tube amps get closer to 'what was recorded'?

The TAG Mclaren ABX test is in the library section; here's a direct link

Also, here is the Stereophile DBT comparison of an SS to tube amp
that I referenced earlier. It's interesting to read not only the results (where a statistically audible difference show evidence of being traceable to a measurable response difference -- the tube amp was more distorting, btw) but the letters that followed (it's amusing to read the messages different readers took home from the article). Note too the large occurence of false positive -- i.e., claiming 'difference' where there was none.


Richard,

H-K has indeed built an expensive speaker comparison facility, described here

 

[Lee Scoggins]

:rolleyes

here we go again...

All tubes do IMHO is make the music more involving and emotional.

...yes and modern tube amps do this by offering a more accurate reproduction of the sound, notably soundstage and midrange.

 

[Chu Gai]

George Orwell would be proud at the redefinition of 'straight wire with gain' as it applies to accuracy. As William Jefferson Clinton said, "It depends on what your definition of 'is' is."

Certainly solid state is capable of generating distortion. As to whether it's audible depends upon the order and magnitude and if the magnitude is small enough, it doesn't matter. It's then not a question of diminishing returns, it's a matter of no returns. Audible that is.

 

[Yogi]

Was I wrong to infer that he feels tube amps get closer to 'what was recorded'?

Yes I think you were wrong in the sense that you took his words literally. I think what he meant (and only Manuel can clear this) was, what I said earlier, that tubes capture the involvement and the emotion of the recording better than SS. It just makes the playback more involving. I think he did discount the fact that tubes might not be as accurate as SS which again IMHO might or not be true. Certianly mating modern tube gear with easy to drive speakers like Von Schweikerts (8 ohm impedance thru most of the FR with a minimum of 7 ohm impedance) has shown a perfectly flat FR just as any good SS amp/speaker combination. So even Chu's often asserted bastardized equalizer theory of tube amps doesn't hold here. So I don't see why SS amps should be any more accurate than a tube amp with properly mated speakers. Tubes might require more work finding the best speaker mate but once the right match has been found the tube amp/speaker system can compete with the best SS amps in terms of accuracy.

Cheers,

 

[Chu Gai]

so we should all go out and VS's?

 

[Yogi]

Chu, by VS's I think you mean Von Schweikerts. That was just an example to illustrate that tube amps with appropriate speakers are as accurate and flat as any SS amp/speaker system. SS amps have a lower output impedance so they are easy to match with any speaker while with tube amps one needs to find the right match due to the higher output impedance.

Cheers,

 

[Scott_N]

Lee

The Soundstage article on the CJ amps is up. It's in the column "The Candy Store" by Bill Brooks.

 

[Mary M S]

A violinist I know states if an instrument is left untouched for months, it has to be ‘played’ in for the tone to become more malleable and friendly again.

Where would any of you with thoughts about the possiblity of varying degrees of individual auditory ability classify this statement as regards a trained professionals ability to detect slight nuances of change in the sound of an instrument?

 

[Yogi]

Can the violinist prove that in a level matched DBT? Just kidding. I believe that statement, but sooner or later someone will ask you this.

 

[RichardHOS]

When I let a trumpet sit idle for a few weeks, it definitely requires some "playing in" to get the sonic characteristics that I'm used to back. That might have something to do with the tendancy for stuff to grow in there if not continuously cleaned.

Steve_AS:

Also, here is the Stereophile DBT comparison of an SS to tube amp
that I referenced earlier. It's interesting to read not only the results (where a statistically audible difference show evidence of being traceable to a measurable response difference -- the tube amp was more distorting, btw) but the letters that followed (it's amusing to read the messages different readers took home from the article). Note too the large occurence of false positive -- i.e., claiming 'difference' where there was none.

I've questioned Stereophile's test methods and statistical reduction techniques for this experiment in the past. I believe it is fairly obvious that they introduced a bias into the test that could explain the small "statistically significant" difference they recorded. That may not actually be the cause of the difference, but the fact that it could be is enough to throw their results into the garbage bin.

 

[Mary M S]

“I think you underestimate my sneakiness”
The violin question was a trick really.

The violin often fascinates physicists. I wonder why…would not a DBT simply prove if the older Cremonese versions truly have something as yet undocumented?
A professor at Syracause University, Wilfred Cote, studies the microstructure of wood.. I read that experiments he has conducted on wood indicate that the long-chain polymers when stressed by sound waves break. Fluctuations in humidity and temp however to some degree repair the chained polymers in the wood. He vibrated violins continuously for 1500 hours, which decreased the B1 (top plate) body vibrations approx. 25 Hz, (ie: making the test instruments easier to play). After a rest of 3 months the B1 body vibration increased about 15 Hz.

Helmholtz Wave. In the 19th century the German physicist Hermann von Helmholtz showed that when a violin string is bowed, it vibrates in a way that is completely different from the sinusoidal standing waves that are familiar to all physicists. The detailed physics of the way a bow excites a string has been extensively studied by Michael McIntyre and Jim Woodhouse at Cambridge University.
Its interesting to note that when comparing the skill of one soloist to another, the Helmholtz wave is particularly difficult for beginning violinist to maintain.

In an article in which author Colin Gough received the 2001 Science Writing Award for Professionals in Acoustics from the Acoustical Society of America, Gough states:
“such factors must have unconsciously guided the radical redesign of the bridge in the 19th century”
he is referring to the acoustically mathematically modeled effects of the sinusoidal force exerted by the bridge on the top plate. How did these violin builders have the gall to change the bridge without scientific evidence to support a break from traditional method? E-gads! They must have relied solely on their ears!.

Heinrich Dunnwald, searching for the global properties of great violins, measured 10 Italian violins, 10 modern (hand fine-grade) and 10 factory made using an electromagnetic driver on one side of the bridge when resonating between 400 and 600 Hz, the factory class were closer to the Italian than the modern copies. A discrepancy occurred however about 1000 Hz with an over-strong response of the moderns as compared to the factory (weak response about 1000 Hz) which would account for comments by musicians about a certain level of shrillness they felt exsisted when describing the quality of sound of the factory built units.

“Every violin, whether a Stradivarius or the cheapest factory-made copy, has a distinctive "voice" of its own. Just as any musician can immediately recognize the difference between Domingo and Pavarotti singing the same operatic aria, so a skilled violinist can distinguish between different qualities in the sound produced by individual Stradivari or Guarneri violins. The challenge for scientists is to characterize such differences by physical measurements. Indeed, over the last century and a half, many famous physicists have been intrigued by the workings of the violin, with Helmholtz, Savart and Raman all making vital contributions….
In practice it is extremely difficult to distinguish between a particularly fine Stradivarius instrument and an indifferent modern copy on the basis of the measured response alone. The ear is a supreme detection device and the brain is a far more sophisticated analyzer of complex sounds than any system yet developed to assess musical quality measurements give the frequencies of important acoustic resonance’s, they tell us nothing about the way a violin actually vibrates. A powerful technique for investigating such vibrations is called time-averaged interference holography. Bernard Richardson, a physicist at Cardiff University in the UK, has made a number of such studies on the guitar and violin……However, neither traditional craftsmanship nor scientific methods can hope to control the detailed resonant structure of an instrument in the acoustically important range above 1 kHz….. What we need is more research, with high-quality violinists working with psycho-acousticians, scientists and sympathetic violin makers, to make further progress in solving this challenging and fascinating problem.” Colin Gough “Science and the Stradivarius” Feature April 2000


My question for all debaters on subjects surrounding [cold Vs warm] [tube Vs SS] [golden ears Vs damaged and challenged by just one too many Rock concerts]
Is Science really the great leveler for those of us modern enough to forgo all outdated superstitions, opinions from great soloist’s, and assorted musicians and audiophiles and rely now soley upon current techniques of modern science including the statisticaly correct analysis of double blinds? Have we arrived at an irrefutable moment in modern times that specs = performance for those enlightened and sophisticated enough to relax or eliminate their reliance upon outmoded personal subjective opinion.

Or is Science still the infant constantly evolving in an attempt to account for the marvel that is the human ear?

 

[Chu Gai]

Specs...nope...relevant measurements gets us moving in the right direction. Might it not behoove one Mary to expend a bit of effort to understand just what is known about human hearing? For example, and I pose the question to you, what do you think is the length of time for auditory memory? For how long can you listen and have your brain recall what it is that you heard? Does the violinist have an extraordinary sensitivity to sounds out of the midrange?

 

[Yogi]

The TAG DBT was not at all rigorous. At least the kind of statistical analysis I am used to seeing is much more thorough. They just established some binomial distribution probabilities and presented the inference. No details about the back ground of the test subjects, or the type of music chosen(not all music would be able to point differences between amps) and a lot more information was missing. Plus the amps it was comparing were pretty much in the same league. Most of the DBT advcates here claim no difference exists between Pass and Aiwa amps. I would like to see a rigorous DBT between amps in vastly disparate leagues. And puhlease dont point me to the silly DBT that compared a Pass Aleph amp and a Yamaha amp. That was the silliest garbage about DBT I ever read.

The moral of the story is that people will believe whatever biases they have, and BTW believing that one is beyond all bias is a bias in itself

So beer, anyone?