HiFi Technology Flatlined Last Century

[dsprito]

Interesting how authoritative opinions can be voiced on foras on topics that are rather clearly far from obvious
So ... there have been I think important audio amplifier designers philosophies that advocated limiting the NFB - so your point is it has always been snake oil ?

 

[levimax]

Interesting how authoritative opinions can be voiced on foras on topics that are rather clearly far from obvious
So ... there have been I think important audio amplifier designers philosophies that advocated limiting the NFB - so your point is it has always been snake oil ?

60 years ago, due to limited open loop gain, NFB was limited to ~20 dB. At this level it did suppress lower order harmonics which "masked" highr order harmonics so the higher order harmonics became more promenent. Since higher order harmics "sound worse" than 2nd or 3rd order harmonics some believed NFB was "bad" but is was far from supported by any controlled listening test. It turns out the "solution" is not less NFB but more. Once you increase NFB well beyond 20 dB all harmonics and all distortion is reduced and it lowers output impeadance... basically better in every way. So the anti- NFB idea, while once based on a sliver of partial fact is now completley bebunked when using modern technology and designs.

 

[Cbdb2]

Bruno P. has managed to influence and convince the whole industry the more feedback the better for the audio quality perception which is controversial to put it mildly.

Only controversial if you you cant understand
https://www.diyaudio.com › community › attachments › the-f-word-bruno-putzeys-pdf.1140285

 

[Cbdb2]

Transistors are very nonlinear devices, with device to device variations and very temperature dependent. Without feed back an amplifiers gain will be different for every channel, change with transistor temp. (which changes with the music), and change with age.
So every amp has feedback, not necessarily global but theres feedback. Theres a myth that only global feedback is a bad thing, The Fword debunks that myth, also.

 

[antcollinet]

60 years ago, due to limited open loop gain, NFB was limited to ~20 dB. At this level it did suppress lower order harmonics which "masked" highr order harmonics so the higher order harmonics became more promenent. Since higher order harmics "sound worse" than 2nd or 3rd order harmonics some believed NFB was "bad" but is was far from supported by any controlled listening test. It turns out the "solution" is not less NFB but more. Once you increase NFB well beyond 20 dB all harmonics and all distortion is reduced and it lowers output impeadance... basically better in every way. So the anti- NFB idea, while once based on a sliver of partial fact is now completley bebunked when using modern technology and designs.

This - thanks for saving me the effort

 

[dguillor]

I could quibble on a few small details but I think you mostly hit the nail on the head.
I wish the truth wasn't so, if only we had something to look forward to equal to the digital revolution of the 1980s.


You're completely correct here boss, but they don't sound any better.

They don’t sound better except that class D can provide lots of power for less price which definitely help the fidelity in some cases.

 

[dsprito]

For me it is not about quoting or recaping some selective materials by fellow audio enthusiasts I am not going to take any technical position on this topic but I can see there are different POV-s on this topic. And they seem to have some ground. Well, at least for me, seems maybe the only one on this forum
E.g. some of the points here

A Future Without Feedback?

Over the years as a reviewer, I have tracked the swings of opinion and popularity of various audio ideas and technologies. Amid a sea of advanced designs that achieve powerful technical performance and laudable specifications, I'm reminded of a major blind listening test of 18 power amplifiers...

www.stereophile.com

do appeal to me. It is 25+ years old but possibly some of the arguments might be grounded in some underlying physics that does not get invalidated by technology progress.
But there have been detailed discussions on NFB I think on this forum so if arguments like in the link above have already been unequivocally and ultimately sentenced as useless there then just let me know and let's not repeat the discussion here in this thread .

 

[mocenigo]

Interesting there has not been a single mention of the digital class D technology which potentially could be the most advanced audio path concept.
Class D now equals analog class D.
Bruno P. has managed to influence and convince the whole industry the more feedback the better for the audio quality perception which is controversial to put it mildly.

Controversial amongst the most illiterate audiophiles. They have no problems with a Lavardin IS just because the manufacturer did state the amount of feedback they used, but it is around 100dB. But, yeah, with Class D it becomes another arrow in the quiver to shoot them.

But what is not controversial it is much more cost efficient to deliver inflated and at some point negligibly relevant frequency domain parameters advantages.

The problem is: how do you correct errors? Because any audio circuit has errors in the analog domain. And at some point any digital representation of the signal must be converted to analog.

Turns out, you cannot avoid feedback there, and for a two-level signal you need a lot of it anyway.

The alternative, which is to use an ADC and feed back to the digital signal, introduces significant delays — and these would mean that any noise and distortion correction would be quite inefficient.

This said, there have been numerous efforts in that direction for ages. Lyngdorf, for example. But they have abandoned it in favour of Eigentakt. Axign and Microchip have their solutions as well. But, they both measure worse AND fare worse also subjectively. So….

 

[mocenigo]

They don’t sound better except that class D can provide lots of power for less price which definitely help the fidelity in some cases.

And lower consumption. So they offer state of the art sound quality (something, you are correct, we have been having for decades) at lower upfront price and lower utility bills. Win-win-win.

 

[mocenigo]

For me it is not about quoting or recaping some selective materials by fellow audio enthusiasts I am not going to take any technical position on this topic but I can see there are different POV-s on this topic. And they seem to have some ground. Well, at least for me, seems maybe the only one on this forum
E.g. some of the points here

A Future Without Feedback?

Over the years as a reviewer, I have tracked the swings of opinion and popularity of various audio ideas and technologies. Amid a sea of advanced designs that achieve powerful technical performance and laudable specifications, I'm reminded of a major blind listening test of 18 power amplifiers...

www.stereophile.com

do appeal to me. It is 25+ years old but possibly some of the arguments might be grounded in some underlying physics that does not get invalidated by technology progress.

The best class D designs of the last 20 years have proved that it is not the amount of the feedback that made sound worse, but HOW it was used.

 

[levimax]

The best class D designs of the last 20 years have proved that it is not the amount of the feedback that made sound worse, but HOW it was used.

It's not just Class D, SOTA Class AB amps like Neurchome use large amounts of feedback to get performance that could only be dreamed of in the past. https://neurochrome.com/products/modulus-86

 

[Old_School_Brad]

And what follows from this understanding ?

Maybe the understanding that negative feedback allows an amplifier to more accurately reproduce the input signal by continuously correcting output deviations, resulting in reduced harmonic and intermodulation distortion, a flatter and extended frequency response, and a lower output impedance that improves damping factor and the ability to drive varying speaker loads.

What do you think it does? -And why is it harmful in your opinion?

 

[Cbdb2]

But there have been detailed discussions on NFB I think on this forum so if arguments like in the link above have already been unequivocally and ultimately sentenced as useless there then just let me know and let's not repeat the discussion here in this thread .

You mean the arguements made by an audiophool who has never designed let alone studied electronics, who writes for a rag that doles out snake oil and subjective BS like theres no tomorrow? What do you think?

 

[Sal1950]

You mean the arguements made by an audiophool who has never designed let alone studied electronics, who writes for a rag that doles out snake oil and subjective BS like theres no tomorrow? What do you think?

Tell them how you really feel. LOL

 

[Cbdb2]

Tell them how you really feel. LOL

Just the facts.

 

[Mart68]

For me it is not about quoting or recaping some selective materials by fellow audio enthusiasts I am not going to take any technical position on this topic but I can see there are different POV-s on this topic. And they seem to have some ground. Well, at least for me, seems maybe the only one on this forum
E.g. some of the points here

A Future Without Feedback?

Over the years as a reviewer, I have tracked the swings of opinion and popularity of various audio ideas and technologies. Amid a sea of advanced designs that achieve powerful technical performance and laudable specifications, I'm reminded of a major blind listening test of 18 power amplifiers...

www.stereophile.com

do appeal to me. It is 25+ years old but possibly some of the arguments might be grounded in some underlying physics that does not get invalidated by technology progress.
But there have been detailed discussions on NFB I think on this forum so if arguments like in the link above have already been unequivocally and ultimately sentenced as useless there then just let me know and let's not repeat the discussion here in this thread .

Article is 25 years old but that preference test referenced was done in 1975.

He presents no grounds for the reason that the Radford being preferred was that it had lower negative feedback. It was the only tube amp in the test. I think it more likely that its interaction with the speaker's impedance was the reason for the preference.

As it happens, I was at a meet earlier this year, listening to some Tannoy Dual Concentrics with solid state amps (Neurochrome I think). A Radford amp (STA15) was swapped in. The loss of fidelity was obvious - but someone remarked that the sound was now easier to listen to - and that was true. Transient hits were softer and bass acquired a slight but not unpleasant 'bloom.'

Nothing to do with negative feedback.

 

[mocenigo]

You mean the arguements made by an audiophool who has never designed let alone studied electronics, who writes for a rag that doles out snake oil and subjective BS like theres no tomorrow? What do you think?

Well, the fact is that they identified an issue with some amps (hard and harsh sound, due to odd order harmonics) and connected it to NFB (correct), and they have drawn the wrong solution, i.e. that better sound could be obtained by removing the NFB. It takes a bit of a leap to understand that the problem was that the hard and harsh sounding amplifiers had too little NFB and that the proper solution was 1) to have more NFB and 2) to shape it depending on the frequency by having filters on the feedback loop — and to guarantee stability even more gymnastics were needed (combine with some feed forward etc...).

I cannot blame them for having reached the wrong conclusion in the first place, and 27 years ago. I blame those that refuse to learn and correct their stance.

Roberto