Feedback in solid state amp


I am pretty sure this topic has been discussed but I am
not sure if this piece of data has been posted.

The general concensus is that feedback lowers distortion
which should mean the sound is better, but based on
listening, no overall feedback amplifier usually sounds
better.

I came across an article from Pass Lab in which he measures
distortion using a single stage FET in class A. What
he found is that with high level of feedback, the overall
distortion is indeed lower vs. with no feedback. But what
is interesting is that although feedback has lower distortion,
there are higher harmonics of distortion that occur in peaks.
These peaks are narrow so they don't add up much to the
overall distortion, but I guess that's why high feedback
amplifier usually sound less musical vs. no feedback amp.

I think similar things occurs in tube amp vs. solid stage
amp, although ss amp has lower distortion vs. tubes, they
have a lot of higher freq. distortion which make them less
musical vs. tubes.

Although I don't have any data, but to some extends, the
same thing happens in low order filter speakers vs. higher
order filter speakers. Low order filter speakers tend
to sound more musical vs. higher order speakers although
higher order filter sounds are clearer.

You can download the Pass Lab article from his website.
The title of the article is:
Audio, Distortion and Feedback
The data the was mentioned above is on page 10.

andy2
The general concensus is that feedback lowers distortion
which should mean the sound is better,

Not just distortion but also improves output impedance. Meaning the amplifier is less controlled by the speaker. 

but based on
listening, no overall feedback amplifier usually sounds
better.

Not universally accepted, like by me for instance, who does not particularly like Pass amps at all. I'll take much less expensive Parasound over Pass any day of the week. They are more conventionally designed. 

Many amps use local feedack instead of global and then even sparingly. Below 0.5% of distortion, it is a wide open playground. 


Although I don't have any data, but to some extends, the
same thing happens in low order filter speakers vs. higher
order filter speakers. Low order filter speakers tend
to sound more musical vs. higher order speakers although
higher order filter sounds are clearer.
 
You are going to have to give specific examples. I don't find this to be true. I prefer Joseph Audio over Thiel and Vandersteen for instance. 

Best,

Erik
My post was not meant for personal preferences.  I guess I won't comment on someone else personal taste.  I am sure there are people prefer solid state over tubes or vice versa.
Erik, he did hit the nail on the head, as explained by the communique by Nelson.

Which part of the distortion is important to human hearing?

engineering weighting does not take any of this into account, which is why it does not correlate to what people hear---- other than in a vague way.

I proposed a good starting point for designing a measurement criteria//method that would have engineering measurements correlate much better to what people hear...over at the DIYAudio forum, in the ’blowtorch’ thread - where some very well respected audio designers tend to hang out.

I’ve yet to witness a result from those communiques, so who knows...