If you are convinced that best amps can be measured, you most likely hit a goldmine of technical wisdom, although the term "smooth" might be not universal enough to define it. However, in my opinion, that is not even possible, especially because amps do not make sound on their own: the sound you hear exists in a relation to all component in the signal path.

I define 'smooth' as a lack of audible higher ordered harmonic distortion as per my prior posts to this thread. The idea that amplifiers do not make a sound of their own is false; we've been hearing these 'sounds' for the last 6 decades. Some amps are smooth and might be a bit soft in the bass, others do bass alright but are harsh in the highs. That this is so is well-known and not controversial.

C’mon folks, apply a little critical reading and thinking skills. Read and see what is clearly between the lines. While I had always admired Ralph and thanked him for his participation here (though disagreeing with his continued insistence that higher cartridge loading is always best unless RFI interferes and that correspondingly, RFI is the chief factor upon which cartridge loading should be chosen), there is clearly a theme here towards promotion of his Class D amps. I have to wonder how a customer of Atma-Sphere who just paid 20K+ for a pair of the MA-1’s a month ago must feel when it’s principal proclaims his own product to be by nature inferior to his new solid state alternative.

And I am also surprised and even angry at myself to only now notice that Ralph is so reliant upon that which can be measured.

:) Critical thinking. I like that: far too little of that going on these days; people have taught themselves to disable their critical thinking. Dietrich Bonhoeffer wrote an interesting paper (which later got him killed) regarding this topic.

Some loudspeakers are not meant to be driven by an amplifier that behaves as a voltage source. ESLs are a good example. For those, tube amps remain the more viable alternative (and just for the record, the last set of MA-1s, which did ship about a month ago, are in use with a set of Sound Labs; something our class D simply can’t make the power to drive because of the impedance. So to my understanding, he feels pretty good.)

Regarding the cartridge RFI thing, that’s easy enough to demonstrate:

http://www.hagtech.com/loading.html

If you have a one meter phono cable, 20pf/foot is typical. find the inductance value of your LOMC cartridge and drop the values into the calculator. You’ll find that the resonant peak is in the RF range. This calculator is based on simple math that comes from electronics 101 in the first week.

Regarding our Class D amplifier, you brought that up, not me. Again, I was addressing the original question, which, again, the answer is that there are semiconductor devices now that didn’t exist 25 years ago, allowing designers to build amps that are actually successful at getting rid of audible distortion (harshness and brightness being an example that might measure quite low on the bench), whereas prior to that time, getting rid of harshness in solid state amps was done with a simple technique called ’lying’.

That lying is why for the last 60 years audiophiles have had to take equipment home to find out if it sounded OK in their system. Its going to take a while to overcome that learned behavior! the first and hardest thing will be to convince people this is all real. As audiophiles we’ve been lied to for so long we take it as rote.

Regarding measurements: One of the problems that many audiophiles have is that the spec sheet seems counterintuitive; amps with lower distortion often sound worse than those with significantly higher distortion. This is because the ear is sensitive to higher ordered harmonics and the harmonics are not ’weighted’ on spec sheets (and the amps with higher distortion are often tube amps with a lot of 2nd and 3rd harmonic). Amps with low distortion figures often have unmasked higher ordered harmonic content at low levels, but the ear is keenly sensitive to these harmonics as it uses them to sense sound pressure, so has a 130dB range to support!

But we can measure what’s important! And reliably predict what the amp will sound like based on those measurements. However most of these measurements rarely if ever show up on spec sheets. Here they are:

1) distortion vs frequency. If rising at all past 1KHz, expect brightness; I explained why earlier.

2) distortion spectra. To mask distortion you have two avenues, either get the distortion product well below -100dB or mask the higher orders with a more prodigious 2nd or 3rd harmonic. SETs use the 2nd harmonic, our OTLs use the 3rd. The distortion spectra should remain consistent at all frequencies, not just 100Hz or 1KHz! FWIW zero feedback tube amps can do this last bit quite well.

3) distortion spectra at -6dB of full power (you never see this one). It should be consistent with the 1 Watt spectra. SETs often fail this test; higher ordered harmonics appear in their output at this level (usually on transients) causing the amplifier to sound ’dynamic’. I’ve found that when audiophiles discuss ’dynamics’, for at least 80% of the time, the word ’distortion’ can be substituted for ’dynamics’ without changing the meaning of the conversation.

If the amplifier is well behaved in these regards it will sound smooth and musical; the lower the distortion the more transparent the amp will be because distortion obscures detail. This is why I don’t like SETs because they often have several orders of magnitude more distortion, causing them to lack detail. Detail does not have to be associated with brightness although it often is because typical solid state amps that are low in distortion are often bright and harsh, although they are at the same time more detailed. You know you’re making progress when the sound is smoother and more detailed both at the same time!

Measurements are really really helpful if the designer understands the how the ear perceives distortion. IOW you can design the amp to work with the human ear rather than to look good on paper; if you really know what you’re doing you can achieve both!

No I meant in a much more simple fashion, i.e. a portion of the output is subtracted from the input. Literally negative. It would technically be out of phase but that is by definition. 

Yes- we use a resistive divider network to do that in our smaller OTLs (we run 2dB of feedback in the M-60). You either do it that way or thru an active device, such as a tube's cathode while the input signal is applied to the grid.

But the feedback network itself can be simple or complex, depending on (...how well its designed...) the expectations for it, such as keeping unwanted frequencies out of the loop. For example in a class D amp that is self-oscillating, the feedback network defines the oscillation frequency (known as the 'oscillation criteria') since so much feedback is applied that the amp goes into oscillation as soon as its turned on; the amp then uses the oscillation as the switching frequency. That network can be a 4th or 5th order loop with 6 or 7 variables; for that you'll need a computer to sort it out.

In addition to what you write, is not the negative in negative feedback, simple because it is subtracted?

If I get your meaning, probably not. In most amplifiers the feedback is applied to a node that has non-linearities, so often feedback signal itself contains distortion not present at the output! For example in a typical solid state amp, the input circuit is a differential amplifier. One of its inputs (differential amplifiers have two, one inverted and the other non-inverted) takes the signal. The other takes the feedback. But the device (transistor) isn't perfectly linear... Of course, if you applied enough feedback, you could overcome this issue...

Let’s simplify negative feedback = taking something ( the AC audio signal aka music ) that has ALREADY happened, flip it around out of phase ( that’s the negative part ) and feed it back into the input where something new and unequal is happening ( unless you think sine ways are music )… then apply some critical thinking….

This statement is false. Its based on the idea that there is a time delay between input and output. What actually happens is there are frequency poles in the amplifier that cause phase shift; on an oscilloscope this appears as a time deviation if you are measuring in the range of the phase shift.

Filter theory tells us that there will be phase shift at some high frequency (in an amplifier, the high frequency roll off will be on a 6dB per octave slop initially; this will impose phase shift to 1/10th the cutoff frequency). As you increase frequency, eventually the phase shift is so acute that the feedback becomes negative instead of positive, and the amp goes into oscillation. Norman Crowhurst was writing about this 65 years ago; none of this is controversial. When the amp goes into oscillation, it can be said that its phase margin has been exceeded.

http://www.tubebooks.org/Books/crowhurst_basic_3.pdf

(link above for those interested in how this really works)

This explains what is really happening and is not some sort of pseudo neo-science.

So if one has been paying attention, one should see what the problem is with feedback in traditional amplifiers (which includes all tube amps). You have an output transformer in most cases, and it causes a fair amount of phase shift, often inside the audio passband! As a result, it poses a limit on how much feedback can be used, and is a guarantee that the feedback will cause harshness and brightness as a result (I already explained why).

Futterman applied 60dB of feedback in his OTLs, but the problem he was up against was not a lot different: his design not only had significant poles in the circuit design, but he also had overall insufficient Gain Bandwidth Product. This means that feedback was decreasing at higher frequencies and so harshness was the result. Because of the frequency poles oscillation was an ever-present danger as well.

Bruno Putzeys wrote a wonderful article on feedback.

https://linearaudio.net/sites/linearaudio.net/files/volume1bp.pdf

It explains much of what I’ve been talking about here in greater detail. Don’t worry if you can’t follow the math 😁 You’ll note that he mentions Peter Baxandall’s writings of how feedback imposes distortion of its own, reflecting Norman Crowhurst’s observations (and explanations) of that from 20 years earlier. Those individuals never got the chance to play with an amp with sufficient bandwidth and loop gain! Its precisely that (sufficient bandwidth and loop gain) which is why an older amp might not keep up with a state of the art design (although it might easily keep up with new amps that are rehashed circuits from earlier decades).

so you are ditching everything you have espoused (if not self-promoted) on this forum for twenty years in order to go all-in to self-promote your new Class D amps.

Wow. Did you see me promoting our class D amps here? I simply stated the underlying engineering issue that any designer faces if they wish to advance the art.

I'm not stating anything in conflict with anything I've stated before. You'll notice that our OTLs are zero feedback- that was done because applying enough feedback to avoid brightness and harshness was never possible. Personally, I find harshness and brightness to be the biggest sin audio can commit. So they are zero feedback, class A(2), a single stage of gain, all triode, fully differential and balanced from input to output, and output transformerless. This was done to eliminate as many distortion sources as possible. It worked- the OTLs have dramatically lower distortion than any SET, are obviously more transparent on that account and have garnered a lot of nice reviews and awards in the high end press. So we know they work, as if our ears weren't telling us that:

Precisely my point , use your ears and decide. If you genuinely believe that the new betters the older alternatives then case closed as far as you’re concerned, I get it.

I would sincerely suggest to the OP to listen to an example of a modern high negative feedback solid state amplifier as described above and compare its sound quality to the Renaissance 70/70 and be your own judge. I’m not convinced that the "modern " high feedback would be  superior sounding.

Emphasis added

I completely agree with this advice, although not his conclusion :)

IME the distortion spectra (the distortion signature) has to be right; too many designers ignore this simple fact.

Most speakers today are designed assuming that the amplifier is able to behave as a voltage source (meaning it can make the same voltage output regardless of load). Some speakers in high end audio are not designed for this behavior- and for those few, sometimes an amp with a high output impedance will sound better.

some modern super low distortion amps can produce a somewhat harmonically lean, clean-white, antiseptic kind of sound that may be not be favored by listeners used to a more saturated, rich, bloomy type of sound that some tube amps deliver...

Once any frequency response issues are sorted, the differences we hear between amps is the distortion signature. You can think of any amp as having a perfect amplifying aspect and also a distortion aspect thru which the signal travels. That distortion aspect is the 'sonic signature' of the amplifier. SETs have a pronounced 2nd and 3rd harmonic, which masks the higher orders (SETs actually have more higher ordered harmonics than any other kind of amp, but when masked you don't hear them), giving them a lush, smooth sound. Some amps which do not have such pronounced 2nd and 3rd have unmasked higher orders, which contribute to the description in the quote above.

That's not just a subjective thing; these aspects are easily measured; and if the proper distortion spectra results in the amplifier design, no matter if solid state or tube, the amp will be easy to listen to, involving and relaxed. So an SET might have the right distortion signature, but imagine two or three orders of magnitude less. you'd hear more detail with no downside (no brightness or harshness).

I know how hard it can be to understand that this is so- for the last 70 years we've simply had to listen to know if an amplifier was going to be musical and satisfying in our systems. That's a lot to overcome! But also for most of that time, building an amp with enough GBP wasn't possible, and the industry really didn't want the market to know that. Heck, that wasn't too hard to sweep under the carpet because of how hard it is to explain what gain bandwidth product even is!

There is always progress.

But to answer your question:

Can a Amp be "timeless" and compete with todays amps?

The correct answer is ’no’. The single thing that has held back amplifier improvement is something called Gain Bandwidth Product (GBP). Over the years you may have noticed that feedback in amplifiers has gotten a bad rap. This is because limitations in GBP have prevented enough feedback from actually being applied. The result has been that the application of feedback has resulted in added distortion on account of the feedback itself, and increasing distortion with frequency above a certain point.

This has caused feedback to be conflated (and rightfully so) with harshness and brightness.

What happens if there is insufficient GBP is that the feedback starts to fall off at higher frequencies and so with less feedback comes greater distortion. If you graph distortion vs frequency you can see this in almost any amplifier made in the last 60 years employing feedback. Since the ear converts distortion into tonality and since the ear is keenly sensitive to the higher ordered harmonics generated by feedback itself, the result is harshness and brightness over the last 60 years.

That is why there are many zero feedback amplifier designs. Zero feedback is about the only traditional way of getting around this problem.

If you can manage about 35 dB of feedback in the amplifier design and there is sufficient GBP, the result is that the feedback can ’clean up’ its own mess, resulting in less higher ordered harmonics. Since the feedback is consistent at all frequencies, the distortion and the distortion signature (the spectra of harmonics) does not change as frequency is increased.

Its a simple fact that no tube amp is capable of this! You need a lot of gain and a lot of bandwidth; the former is simply not available in tube designs without associated phase shift, which would cause the amp to oscillate at some higher frequency.

There are now designs that satisfy these goals, designs that were not possible 20 or 30 years ago because the amplifying devices didn’t exist.

What this means is its possible to build a solid state amp that is every bit as smooth in the mids and highs as the best tube amps and leave nothing on the table in terms of detail, sound stage and the like, in such a way that vintage amps simply cannot compete (not that they sound bad, just they don’t sound as good). This is one of those things that is easy to hear and easy to measure.