Amplifier Feedback Good or Bad?

[highfigh]

The hardware and broadcasting system for XM Radio was far superior to Sirius, the XM system used (3) satellites, and cell tower repeaters. However the primary advantage Sirius was its diversity for content, I have XM/Sirius in 1 of my vehicle but rarely use it but nice to tune in a target music channel without commercials. In concluding the satellite radio audio sonics are better than AM radio but still not audiophile quality....

Just my $0.02...

Remember AM Stereo? That was just.....great.

 

[M Code]

Yup..
C-Quam AM stereo.. Nice try..
Remember it well, I worked very closely with the CEO of Motorola for its introduction into certain products. Interesting idea but little market demand or support....

Just my $0.02....

 

[Auditor55]

The range you posted is the supposed 'normal' hearing acuity range for humans, at some point in their life, but it refers to the physical stimulus of our hearing mechanism, not how we perceive the sound and that's where psychoacoustics comes in. Can we localize all frequencies as coming from the speakers? No, because as I wrote, some frequencies seem to be coming from someplace that would be impossible even when the listener/patient is wearing headphones.

Psychoacoustics is the branch of psycho-physics involving the scientific study of sound perception and audiology—how humans perceive various sounds.

It all falls under the field or study of Psychoacoustics . The supposed arrange of normal human hearing was determined by what method? I would say that it was determined by listening test (of actual human beings), to which such test had objective results, of which were that average human being could only hear sounds between 20hz-20khz.

 

[Auditor55]

Anyway, why should amplifiers add any coloration to the sound?

 

[highfigh]

Psychoacoustics is the branch of psycho-physics involving the scientific study of sound perception and audiology—how humans perceive various sounds.

It all falls under the field or study of Psychoacoustics . The supposed arrange of normal human hearing was determined by what method? I would say that it was determined by listening test (of actual human beings), to which such test had objective results, of which were that average human being could only hear sounds between 20hz-20khz.

Basic hearing tests only check to determine IF the person can hear the sound, not where it seems to come from, its timbre, whether it's irritating, etc. Look into the Fletcher-Munson Equal Loudness Contour. In a basic test, they only go to 8KHz because they're concerned with the person's ability to hear the range needed for speech, not the full range or for music enjoyment.

 

[mtrycrafts]

Anyway, why should amplifiers add any coloration to the sound?

Well, do you just look at a blank canvas or all painted up.

Of course audio should be colorless.

 

[Marcus Fry]

Aren't Ayre amps one of those makers who stays away from feedback, or at least used to?

 

[Gmoney]

Well, do you just look at a blank canvas or all painted up.

Of course audio should be colorless.

amps are not Allowed to color, only tone controls or EQ or DSP that there job.

 

[mtrycrafts]

amps are not Allowed to color, only tone controls or EQ or DSP that there job.

One would think.

 

[Gmoney]

One would think.

myrycrafts, most don’t use any of those controls some do I use some it depends on the Source or the recorded source. I have picked up CDs that where lifeless. or a movie mostly the ones made in Mexico you know the ones you can pickup at Walmart for 5 bucks. Not saying Walmart sells bad discs sometimes I pickup a gem like that Assassin move Wanted, with Morgan Freeman and Angelina Jolie. 5 buck movie 5.1 nice sound track.

 

[Gmoney]

@ Gene, Have you ever done an Article on Dampening factor, Amplifier Dampening Factor Good or Bad: Than pole up, Sounds better with more or less. Some amp Manufacturers do postup specs of Dampening Factor.

 

[Gmoney]

Me neither, I am not an audio electronic engineer, though I am sure if there is such a degree option. My final year options for my EE degree was in Control and power system. "Control systems" means the syllabus would include feeback, feed forward theories, and lots of Laplace, Z, Fourier transforms, infinite series kind of math. Just because someone is an EE, doesn't mean one has in depth knowledge in audio electronics, though one usually would have/or should have the basic knowledge.

Getting on the Subject of degrees , last 7 year’s in the petroleum industry went to Technical schools Controls and Safety systems, Operations of PCL and repair of Electrical and pneumatics End Device also MWCP Shut down systems and Communications systems Data systems Installation of and troubleshooting All.
You wouldn’t believe some of the oldest Technology that still being used in petroleum industry on Production platforms. Outdated unsafe and a down right shame for the lack of regard for the Safety of the environment and human life. Some of those OP out there are Contracted about all some of them can do is keep the place up and running but have No working knowledge of why those Safety systems are put into place. Oh many have Learned from on the job years of use or by placing calls in and around the Blocks out there. Think some old retired person who has 30 + years they will Hire a guy like that as a safety consultant or consulted. Think New Horizon explosion in the Gulf of Mexico. He is Usually paid by the Oil company or sent out as Independent. I Firmly believe that on any Drilling should have a authority from MMS On board at all times. Think like a Federal agent was they are supposed to be anyways During Operations until completed and the downhole is plugged. It is in my Opinion that No Venue of Safety should be over looked because of cost or lack of. But than at A Human Life is the Cheapest thing to replace it the Oilfield.

 

[Bruno P.]

Aren't Ayre amps one of those makers who stays away from feedback, or at least used to?

Yes. Which is funny because Charlie and I were very much agreed on most other things. In particular the crazy constant race towards new digital formats and ever higher speeds which is just a ludicrous distraction. We were very much on the same side there. The expectation that significantly better sound can be had by storing ever more bits for a second of audio is just baffling for anyone who's ever heard an ADC/DAC chain switched in and out of a monitor loop with barely an audible change. The only sensible explanation then is that the people who advocate high sampling rates simply haven't worked out how to do digital band-limiting filters that you can't hear. In other words, super high sampling rates mostly benefit badly designed DACs.

I can't help but see a parallel with feedback denialism though. There too, I've yet to run into someone with solid grounding in control theory who then goes on to say "but I avoid it in audio amplifiers". That just never happens. The honest thing for designers of non-feedback amps would be to say: "well feedback ain't my strongest suit, so here's the best I did without it. Hope you like it." If they did that, I'd probably even say "great, let's hear it".

 

[Gmoney]

audiocorr

Yes. Which is funny because Charlie and I were very much agreed on most other things. In particular the crazy constant race towards new digital formats and ever higher speeds which is just a ludicrous distraction. We were very much on the same side there. The expectation that significantly better sound can be had by storing ever more bits for a second of audio is just baffling for anyone who's ever heard an ADC/DAC chain switched in and out of a monitor loop with barely an audible change. The only sensible explanation then is that the people who advocate high sampling rates simply haven't worked out how to do digital band-limiting filters that you can't hear. In other words, super high sampling rates mostly benefit badly designed DACs.

I can't help but see a parallel with feedback denialism though. There too, I've yet to run into someone with solid grounding in control theory who then goes on to say "but I avoid it in audio amplifiers". That just never happens. The honest thing for designers of non-feedback amps would be to say: "well feedback ain't my strongest suit, so here's the best I did without it. Hope you like it." If they did that, I'd probably even say "great, let's hear it".

@bruno P, Those Pro Saber ESS DAC that Yamaha use’s in their Flagship AVR a3080 and their 5100 and 5200 are they any better than what Denon or Onkyo or Marantz use? Also the Burr-Brown. Your thoughts on those. Your short Yes? No? Your post was a Interesting read.

 

[highfigh]

I can't help but see a parallel with feedback denialism though. There too, I've yet to run into someone with solid grounding in control theory who then goes on to say "but I avoid it in audio amplifiers". That just never happens. The honest thing for designers of non-feedback amps would be to say: "well feedback ain't my strongest suit, so here's the best I did without it. Hope you like it." If they did that, I'd probably even say "great, let's hear it".

I want to buy an amp with zero NFB, then connect some oddball speakers, goofy-looking power cords, speaker wire and interconnects with supposed magical properties and put everything in a room with bad acoustics so I can invite some hard core audiopiles for a listening session, complete with handouts showing the lack of NFB in the amp and creatively written descriptions of what they'll hear.

I expect the reviews to be glowing.

 

[PENG]

The only sensible explanation then is that the people who advocate high sampling rates simply haven't worked out how to do digital band-limiting filters that you can't hear. In other words, super high sampling rates mostly benefit badly designed DACs.

I think you are referring to designers/engineers. As an end user, I would definitely recommend people focus on high sampling rate and bit depth files from the likes of hdtracks.com not because that makes them sound better but because in my experience high sampling rate/bit depth including DSD files seem to have a greater chance that the recording/mastering quality are better than those at lower rates such as 44.1/16 bit files. I don't know why that is, but that's unfortunately what I experienced so I am willing to pay more for those.

I can't help but see a parallel with feedback denialism though. There too, I've yet to run into someone with solid grounding in control theory who then goes on to say "but I avoid it in audio amplifiers". That just never happens. The honest thing for designers of non-feedback amps would be to say: "well feedback ain't my strongest suit, so here's the best I did without it. Hope you like it." If they did that, I'd probably even say "great, let's hear it".

I don't know about you Bruno but I am sure you are much smarter than I am and probably a lot of so called designers/engineers too. When I was doing my EE degree with control theory as one of the two options I found control systems theory, like communication theory, were very difficult to understand in terms of their basic concepts and practical values, as well as the advanced math involved. I was quite okay with the math, but the concepts of Fourier's, and the other transforms (Laplace, Z.. etc.,) could be hard to grasp for a lot of students, let alone their applications. It was not until years later when I was forced to take a course in communications that I understood (thought I did then, at least) what Fourier series/transform is about. Prior to that, I had no clue how and why an ugly waveform can be treated as a series of sine waves, though that's supposed to be so basic! It is like Calculus, how many people don't realize the basic idea may involve the concept of approximation (delta..), but the results are exact..

So all I am trying to say is that it may be possible that many engineers/designers don't really understand the concepts related to the feedback and control theories and are therefore not too comfortable in using it and/or relying on it when they apply it in real world products such as amplifiers. Though I am sure many wouldn't be concerned about auto pilot on a jetliner, or even pay to land on the moon someday, lol..

It could also be that some one like your smart friend, who actually know, and agree with you on the use of feedback without the perceived inherent negative effects on sound quality, but they have their products to sell. I think the highly regarded audio electronics like Mr. Pass, and Dr. Geddes Gedlee might now have different, or softened views on things like the much hyped negative impacts on "sound quality" of the so called excessive feedback, TIM due to poor slew rate and the once much worried about DF etc etc.., but I highly doubt any of them would publish new papers on their softened or even changed view, even if they did have one.., especially if they still have products to sell that has been benefited by the popular beliefs among "audiophiles" and the hearsay prone "audioholics", or just beginner hobbyists.

By the way, it is great to have you posting on Audioholics, we are lucky to have experts like you to discuss things with. After building the simple 5 W Nelson Pass Class A amp (obvious very little feedback used on that one), I am quite sure my next project will be one of those Hypex Ncore amps.

 

[mtrycrafts]

@ Gene, Have you ever done an Article on Dampening factor, Amplifier Dampening Factor Good or Bad: Than pole up, Sounds better with more or less. Some amp Manufacturers do postup specs of Dampening Factor.

I am pretty sure he has something on it. If not, check this out:

Redirecting to Google Groups

It is a raw number amp output impedance and speaker impedance at frequencies.
When output impedance is high, like some tube amps, DP can be rather low, 8 or less. There are better out there.

 

[Bruno P.]

Those Pro Saber ESS DAC (...) are they any better than what (...)

Thread hijack risk. But to give you a short answer, I've got two ways in which I design DACs. One is from the ground up when I need to do an absolute showpiece like the Mola-Mola. For that purpose I don't go looking at what DAC chips are available because I shan't be needing them. The other is pragmatic, "I need a good enough analogue output with minimal effort", in which case I default to AKM because I've never had any nasty surprises with them. In that case though I am ready to accept compromises like greater-than-necessary audible changes between sampling rates etc. Either way I don't really go hunting for the chip du jour and neither do I open up other folks' gear to read the print on the parts they use.

I would definitely recommend people focus on high sampling rate and bit depth files from the likes of hdtracks.com (...) because in my experience high sampling rate/bit depth including DSD files seem to have a greater chance that the recording/mastering quality are better than those at lower rates such as 44.1/16 bit files.

Yeppers. That's also my answer when people press me about sonic differences with specific examples. A recording engineer who worries about digital formats is also likely to agonise over mic choice and placement, processing and (let's be optimistic) arrangement and orchestration.

I found control systems theory, like communication theory, were very difficult to understand in terms of their basic concepts and practical values, as well as the advanced math involved. I was quite okay with the math, but the concepts of Fourier's, and the other transforms (Laplace, Z.. etc.,) could be hard to grasp for a lot of students, let alone their applications.
(...) it may be possible that many engineers/designers don't really understand the concepts related to the feedback and control theories and are therefore not too comfortable in using it.

I think that's absolutely fair. I don't blame anyone for not understanding feedback theory. I don't blame them for wanting to build and sell amplifiers notwithstanding either. What I think is odious is their covering up for that lacuna by claiming to understand it so well that they just know it's bad. Giving up and declaring victory is what that is.

Let's face it. We all slept through control theory classes. I also had to teach myself on the job even though it was on the curriculum. The difference between me and, say, the other Mr. P is that I decided that it was important to figure out about control theory before making sweeping statements about its usefulness in audio. That was helped by the fact that I also wanted to learn the fundamentals of sigma-delta AD/DA conversion and that was all about feedback. And then I realised that by comparison amplifier designers had only scratched the surface and come to a rather hasty conclusion based on very incomplete experience.

By the way, it is great to have you posting on Audioholics, we are lucky to have experts like you to discuss things with.

In which case I must apologise in advance that I won't be terribly active on threads other than those linked specifically to Audioholics editorials and talks I take part in.

I am quite sure my next project will be one of those Hypex Ncore amps.

Thanks! Mind you, I left Hypex in 2015 and (shameless plug alert!) I actually have a new generation of amps out with a new company called Purifi. Search for "Eigentakt".

 

[Gmoney]

Thread hijack risk. But to give you a short answer, I've got two ways in which I design DACs. One is from the ground up when I need to do an absolute showpiece like the Mola-Mola. For that purpose I don't go looking at what DAC chips are available because I shan't be needing them. The other is pragmatic, "I need a good enough analogue output with minimal effort", in which case I default to AKM because I've never had any nasty surprises with them. In that case though I am ready to accept compromises like greater-than-necessary audible changes between sampling rates etc. Either way I don't really go hunting for the chip du jour and neither do I open up other folks' gear to read the print on the parts they use.

Yeppers. That's also my answer when people press me about sonic differences with specific examples. A recording engineer who worries about digital formats is also likely to agonise over mic choice and placement, processing and (let's be optimistic) arrangement and orchestration.

I think that's absolutely fair. I don't blame anyone for not understanding feedback theory. I don't blame them for wanting to build and sell amplifiers notwithstanding either. What I think is odious is their covering up for that lacuna by claiming to understand it so well that they just know it's bad. Giving up and declaring victory is what that is.

Let's face it. We all slept through control theory classes. I also had to teach myself on the job even though it was on the curriculum. The difference between me and, say, the other Mr. P is that I decided that it was important to figure out about control theory before making sweeping statements about its usefulness in audio. That was helped by the fact that I also wanted to learn the fundamentals of sigma-delta AD/DA conversion and that was all about feedback. And then I realised that by comparison amplifier designers had only scratched the surface and come to a rather hasty conclusion based on very incomplete experience.

In which case I must apologise in advance that I won't be terribly active on threads other than those linked specifically to Audioholics editorials and talks I take part in.

Thanks! Mind you, I left Hypex in 2015 and (shameless plug alert!) I actually have a new generation of amps out with a new company called Purifi. Search for "Eigentakt".

Thank you Bruno P, coming from One who Actually Works, builds and Designs audio gear is Refreshing.
I have always stated that it all begins at the pre-amp side. crap in crap out. Before I met you on AH, all us Audioholic junks, really has no one other than researching. PENG stated in one of his post that the one’s take Manufacturers Marketing post of specs and so on at face value and PENG in my Opinion has been more correct on AH than most who just jump in and post up specs. Gene can’t alway jump in every Thread or post very busy guy but he has at times.

 

[Pogre]

I listened to this video last week while I was driving so I wasn't watching. Every time Bruno spoke I kept picturing Richard Dawkins. Bruno sounds just like him!