The imperfect amp: Pass or Ayre?

IMO THD measurements are among the most useless of specs, and I would by no means consider amps having relatively high THD numbers, compared to competitive products, to be ones that "from a technical perspective, don’t do very well" (quoting from the OP).

It’s pretty well established that while large amounts of feedback can reduce THD to very small amounts, that usually comes with two significant side-effects:

1) Increased amounts of Transient Intermodulation Distortion, which is not normally measured, and which as far as I am aware does not even have a standardized basis for measurement. That despite the fact that its significance was recognized as far back as the 1970s, when Dr. Matti Otala famously authored several papers on the subject, this being one example:

https://pdfs.semanticscholar.org/b3c0/a892a982ebde91f83f228905dac30186f827.pdf

2) Increases in some higher order odd harmonic distortion components, which occur even though Total Harmonic Distortion is reduced. As Ralph (Atmasphere) has pointed out many times, our hearing mechanisms are extraordinarily sensitive to certain higher order distortion components.

As Gregm aptly quoted above:

If something sounds good & measures well, it is good; if it sounds good and measures bad, you’re measuring the wrong thing!

I believe, btw, that was originally stated ca. 1960 by Daniel von Recklinghausen when he was the chief engineer of the original H. H. Scott company. He later worked for KLH, among other firms, and served for a while as president of the Audio Engineering Society.

Wolf_Garcia, LOL! Thanks for another example of your inimitable humor.

Regards,
-- Al

It’s advertising BS by Ayre to say "no Feedback!" on a solid state amp.

This statement is false. Although there are not many examples, it is possible if build a solid state amp that lacks any sort of feedback whatsoever.

Both Pass and Ayre are two amps which eschew technical perfection.


Both Ayre and Pass stand in contrast to say, the Halcro amplifiers, which were famous for their absolute dominance in technical terms. Ayre and Pass however have ended with different followers.

@erik_squires 
IMO/IME you've got this backwards. Ayre and Pass are great successes, because they sorted out what is important to the human ear and applied engineering to design a circuit that more closely obeyed the rules of human hearing/perception. This was done in the face of a test and measurement regime that ignores how the ear works and instead concentrates on measurements that have little to do with how the ear perceives sound.

Now if you are coming from the perspective of 'objective measurement' this will come off as a bit of heresy. But the simple fact of the matter is 'objective measurement' isn't objective; at best appears to be a mindset that just gets really riled up when things don't go its way. To this end if an amplifier circuit is designed to sound good to the human ear, it gets lambasted for 'high distortion' in the face of what we know at this time of how the ear perceives distortion. 

So a bit of a primer: lower orders (2nd, 3rd and 4th) are perceived as 'richness', 'bloom', 'warmth' etc. But they do more than that, but first its important to understand that the ear perceives the higher ordered harmonics as harshness and brightness- in particular the odd orders. Now the ear also has a masking principle- how a louder sound masks the presence of a quieter sound. In the case of distortion, the lower orders are useful in masking the higher orders. This is important as the ear uses higher ordered harmonics to sense sound pressure (which is very easy to demonstrate with simple test equipment).

So imagine what happens when feedback suppresses the lower orders (and for that matter, the higher orders too) but in the process **adds distortion of its own** through a process called 'bifurcation' (which, not by coincidence, is a common term to those familiar with Chaos Theory). Also not coincidental is the fact that some basic Chaotic systems have mathematical formulae that are strikingly similar to the feedback formula!

So in an amplifier that employs feedback and thus has low THD, the remaining THD that is *does* have is mostly there from the operation of the feedback itself (again this phenomena is well-known; see the writings of Norman Crowhurst from 60 years ago for more). So this distortion, while low, has nothing to mask it (the lower orders are suppressed) and so easily stands out to the human ear as brightness and harshness. Some people describe it as 'crisp'.


Put another way, an amp that has higher levels of higher ordered harmonic distortion will sound smoother and not as bright **if** it also has lower orders (in particular the 2nd or 3rd) to mask the higher orders. This is why tube amps have higher overall distortion across the entire spectrum yet sound smoother and not as bright; this having nothing at all to do with bandwidth, since the ear is converting that distortion into a tonality. In a sense the tubes are very good at obeying human hearing perceptual rules- that is why they are still around.


So neither Nelson or Charlie were/are going for a tube sound so much as they are going for a **natural** sound. The problem is simple to any pragmatic designer: its impossible to build a circuit with zero distortion! So next best isn't to make an amp that is as low distortion as possible, the next best is to make an amplifier that has the lowest possible **perceived** distortion, which is something very different. Since the ear is relatively insensitive to the lower orders. such a design makes use of that fact and the masking principle to create an amplifier that actually sounds neutral (IOW, more like real music).


Something peculiar here is that amps with a prominent 2nd or 3rd harmonic also are perceived as more detailed with a more palpable sound stage- better center-fill. Why this is so isn't well understood and might be an excellent topic of inquiry.


So you've singled out the two most commercially successful solid state amps that address these issues!. They aren't failures!- instead they are **leaders** which will be historically recognized as such when the rest of the industry sorts the above facts about human hearing. But I'm not holding my breath on that because the spec sheets are some of the best advertising any inferior amp manufacturer has at their disposal; 'see?- very low distortion- it has to be great!'; an excellent example of the Emperor's New Clothes. If things were otherwise we could simply look at the spec sheet and know how the amp sounds and we wouldn't have to audition it in our systems to know it would work. In a nutshell, the industry has been lying for at least the last 60 years- and we've all grown up with that so don't see it as weird that despite the spec sheets we still have to do an audition.

@atmasphere

Thanks for that *very* informative post! I think it's a telling anecdote that Halcro is no longer a thing (THE BEST AMPLIFIER EVER! - lol).

How many other pursuits in life are people so obsessed with measurements in a realm where the ultimate goal is to engage in something pleasing to our senses? I won't possibly like the way this tastes because of the absence of sugar or salt? This artwork doesn't move me because it contains too much in the blue-green spectrum?  If it sounds good enjoy.  Why sabotage your happiness by questioning whether you truly like it while critically assessing measurements that perhaps don't seem to quantitatively equal perfect sound?

My power amp, one of Dennis Had's SEP High Output tubsters (10 to 17 WPC allegedly), is one that I hope nobody ever measures for anything as that could harsh my mellow.