What is meant exactly by the description 'more musical'?

Interesting , some people can hear negative feedback = TIM and simple temporal distortion in addition to the various forms of harmonic distortion Ralph and a FEW ear/brain centric designers ( very rightly but not completely ) focus on.

@tomic601 Timing usually isn’t a problem with negative feedback. Phase shift is (and can look like a timing problem if you’re only using a ’scope to see what’s going on). TIM really isn’t a problem with feedback so much as it was a problem associated with part of the (high feedback, solid state) amplifier that was outside the feedback loop: the base of the input transistor or perhaps the entire transistor, depending on the design... which would easily distort on transients.

Norman Crowhurst pointed to the problem of a non-linear feedback point (node) in an amplifier, such as the cathode of the input tube 65 years ago. He points out that due to that non-linearity, the feedback causes a noise floor of higher ordered harmonics as well as inharmonic (intermodulation) distortions. He did not propose a fix. Later, Peter Baxandall pointed to the same problem in solid state amps and simply proposed more feedback as a fix, which does not work; any signal distorted by that non-linearity (in this case, the base of a transistor) will not somehow magically be healed by more feedback.

The solution is to mix the feedback with the incoming signal using a resistor divider network, before the signal gets to the input of the amp (the way opamps do it). In that way the feedback signal is not distorted and so can do its job more effectively. FWIW this is how we’ve applied feedback in our OTLs for the last 35 years or so.

Since most people believed Baxandall and don’t remember Crowhurst talking about the same problem 20 years earlier, we wound up with a lot of amps that were less than musical in the last 70 years. The issue of increasing distortion with frequency really didn’t show up for decades after! So now you have something to look for in the measurements that relates directly to that mysterious ’musical’ aspect some amps have and some don’t.

For those who want to dig into more details about this feed back problem which atmasphere explain just above :

Read the first 24 pages article titled

The sneaky pitfalls of feedback and feedback theory

https://www.temporalcoherence.nl/cms/en/pyra

It will give you a gist about the complexities of the problem...

I myself cannot make any wise observation about amplifier design at all ..

We are lucky to have atmasphere here regularly for answering questions ...

Ralph, I have to believe that musicality, immersion, involvement and, ultimately, suspension of disbelief has much to do with an amp designer’s taste and ear than simply harmonics and distortion.

I would say that your sensibilities as a musician have as much, or more, to do with your amp’s andpreamp’s sound as your technical abilities.

The sneaky pitfall is taking a waveform that has already happened, reducing it, flipping it out of phase and feeding it back and expecting it to heal a different waveform. The issue is can the ear brain perceive and evaluate this as “ musical “.

One reason why many but not all low to no negative FB designers use time and phase accurate speakers… Always cracks me up when phase accuracy obsession in electronics is swamped by higher order filters downstream.

I do have a lot of respect for Ralph, and i would say his gear is very musical. There are of course others working the same problems w different perspectives… and some probably now know the ANSWERS…thinking in particular of Roger and Charlie ( Ayre )

Best to you Ralph !

@mglik You are right but everyone uses the same hearing perceptual rules. Its a sort of common denominator.

Over time, as a designer you gain experience knowing how the distortion profile looks as compared to how it sounds. It gets to the point that you can predict how the equipment will sound if you have enough of the relevant measurements in front of you. 

There is that old saw about trust your ears not the measurements. That was really true until sometime in the late 1980s or early 1990s. Sometime in that period it became possible to measure the things that really tell what an amplifier (or preamp) might sound like. The knowledge that this is so is scarce, as is the understanding of the significance of some of the measurements.

At any rate, if the design issues I mentioned so far are not observed in a design, it is very likely it will not be deemed 'musical'. It might surprise you, but we've never relied on tuning any of our products by ear. We rely very heavily on measurements. Only after we get what we are looking for do we give it a listen. So far this technique has served us pretty well.