No - not in the sense as it was asked for.
I have proven your point of view to be wrong by my measurements over the course of this thread - and Bob is pointing out the same from the principle of operation .
So, better update your "science", Earl... 😉
Michael
the line by line "point scoring" deconstruction/semantic analysis is really wasting time and annoying to the subject and viewers as well
in an informal exchange accept that shortcuts, approximations and desire for speedy communication may lead to statements that are “wrong” in an academic paper or textbook setting but can actually aid information exchange if sophisticated participants look at the “big picture” argument
and due to the wide reach of the web different educational backgrounds will have us all not immediately understanding or liking the nuance or presentation of even things we do recognize and (mostly) agree with
why not accept that Dr Geddes is a genuine expert and try to use your own expertise to construe any semantic niggling in a favorable direction rather than being needlessly argumentative
asking for clarification of central argument points is often necessary - but try to limit it to the central point - we'll all learn a lot more, faster that way
in an informal exchange accept that shortcuts, approximations and desire for speedy communication may lead to statements that are “wrong” in an academic paper or textbook setting but can actually aid information exchange if sophisticated participants look at the “big picture” argument
and due to the wide reach of the web different educational backgrounds will have us all not immediately understanding or liking the nuance or presentation of even things we do recognize and (mostly) agree with
why not accept that Dr Geddes is a genuine expert and try to use your own expertise to construe any semantic niggling in a favorable direction rather than being needlessly argumentative
asking for clarification of central argument points is often necessary - but try to limit it to the central point - we'll all learn a lot more, faster that way
And the summarised 'analysis' is not?😛 He's making some fairly basic errors and not accepting he's wrong. If you feel annoyed by attention to detail, feel free to move along, there's no compulsion to read.
I accept he's a genuine expert in his field of acoustics, clearly electronics is not his area of expertise to the same extent. But this is merely ad hominem, and hence a waste of time - a man's arguments stand or fall by the strength of his arguments, not any 'expertise' he may claim to have or not have.
telling me "feel free to move along" - is exactly the abrasive point scoring argumentative style that we really don't want to encourage here
Last edited:
How good is as good as perfect?
Referring to your paper "Auditory Perception of Nonlinear Distortion", if I believed that the regression line on Figure 5 was the absolutely correct measure, then any Gm below 0.75 would be perfect, because the line meets 0 (horizontal axis) at Gm=0.75. On the other hand, using the few dots above the regression line I could try to draw a new line that meets zero at approximately Gm=0.25 . Now, if someone built an amplifier with Gm under 0.25 with a safe margin, it should sound good. I am talking about otherwise perfect amplifier without any other distortions depending on frequency or feedback.
Am I interpreting too much? Perhaps only perfect is really good.
Referring to your paper "Auditory Perception of Nonlinear Distortion", if I believed that the regression line on Figure 5 was the absolutely correct measure, then any Gm below 0.75 would be perfect, because the line meets 0 (horizontal axis) at Gm=0.75. On the other hand, using the few dots above the regression line I could try to draw a new line that meets zero at approximately Gm=0.25 . Now, if someone built an amplifier with Gm under 0.25 with a safe margin, it should sound good. I am talking about otherwise perfect amplifier without any other distortions depending on frequency or feedback.
Am I interpreting too much? Perhaps only perfect is really good.
Statistically, based on our data, if the value is below 1.0 then it is "indistinguishable" form "inaudible". Or said another way, if Gm < 1.0 then it is not significantly different than "inaudible". I would call this "good enough". There could be the exceptional situation where < 1.0 was audible, but 99% of the time it won't be. I would even be tempted to use a value of 2.0 as "good enough". Clearly, somewhere between 1.0 and 2.0 is certainly the point that you are looking for I believe.
Hilarious😀
The postings that I really don't want to encourage here are ones where
1) the poster snips words out of context to prove a point they themself already believe
2) the poster besmirches the style of another person's posting
3) the poster fails to address the meat of the other person's arguments
4) the poster claims support from others for their point without offering evidence
I'm on to the second of your AES conference papers and found the following:
It is well known that an MP3 sound
transmission can have a measured THD of upwards
of 50%, and yet be perceived by listeners as an
acceptable quality reproduction.
You don't cite a reference for this claim. Do you have one?
A buyer's guide?
Dear Sirs
I'd really like to know, what have we learned from this about all the ordinary, everyday amplifiers sold in kitchen appliance stores?
I have in my bookself a NAD C325 BEE integrated amplifier. I keep wondering if I should buy a second hand Pioneer VSX-D912 to be sure I'm not missing on anything.
I have an EMU 0404USB module that is supposed to have 24-bit sound output. It is connected to a Mac Mini and I can run linux and different versions of Windows in a virtual machine. I wonder if the secrets of the subject "came out" already or not, and if they did, is there an easy way for me to measure my NAD integrated amp?
I wish a magazine would 1) adopt this way of measuring, and 2) find out when and how the findings correlate with the opinions of people with the so called golden ears.
But what would you say? Am I safe with my NAD C325 BEE integrated and C275 BEE power amplifier? Should I buy a second-hand VSX-D912 or is there even something better available off the shelf? I have the service manuals for VSX-D912 and VSX-920, and the latter uses discrete parts where the former uses hybrid IC chips. A local dealer said that D912 is a discontinued model from year 2002, replaced by VSX-D918 in 2008.
Dear Sirs
I'd really like to know, what have we learned from this about all the ordinary, everyday amplifiers sold in kitchen appliance stores?
I have in my bookself a NAD C325 BEE integrated amplifier. I keep wondering if I should buy a second hand Pioneer VSX-D912 to be sure I'm not missing on anything.
I have an EMU 0404USB module that is supposed to have 24-bit sound output. It is connected to a Mac Mini and I can run linux and different versions of Windows in a virtual machine. I wonder if the secrets of the subject "came out" already or not, and if they did, is there an easy way for me to measure my NAD integrated amp?
I wish a magazine would 1) adopt this way of measuring, and 2) find out when and how the findings correlate with the opinions of people with the so called golden ears.
But what would you say? Am I safe with my NAD C325 BEE integrated and C275 BEE power amplifier? Should I buy a second-hand VSX-D912 or is there even something better available off the shelf? I have the service manuals for VSX-D912 and VSX-920, and the latter uses discrete parts where the former uses hybrid IC chips. A local dealer said that D912 is a discontinued model from year 2002, replaced by VSX-D918 in 2008.
First, the opinions of "Golden Ears" are about useless since they consistantly hear things that don't exist and cannot be heard in blind tests by them or anyone else. But yes, there are things in amplifiers that matter. Its mostly the nonlinearity at very low signal levels - something that no one publishes. So test your amps for the harmonic content as you reduce the signal level. The harmonics of a good amp will disapear into the noise. A bad map will have the higher orders actually grow in relative level as the signal level drops.
I don't speak to specific amps, but many Hybrid amps work very well (its a lot easier to do a good amp if everything is on a single substrate - the components all thermally track better - better matching means better crossover transition.) I have found almost no colleation with brand, price, or topology with good sound quality (as defined by measurements).
Concur as far as opinions go, but the golden eared boys do sometimes describe what they're hearing. Those descriptions are useful to me as I'm keen to understand what electronic artifacts might be responsible for their perceptions.
In my experience (not my opinion) is not mostly that (haven't encountered amps with low level distortion problems, though I'm not denying they exist). Its mostly the RF immunity of the amp which matters, which is strongly related to filtering and grounding issues. In my opinion this is one reason why golden eared types tend to prefer valve amps - they don't have low level distortion (but have bucketloads of higher level, low order stuff) and they generally have good RF immunity.
Yep, that's a useful test, but requires FFT analysis. One test you can do easily is test the common mode rejection of an amp - see what comes out of it when just the ground of an input is connected to an RF noise source ( the common/ground terminal of a cheap DVD player which has an SMPSU is one possible source, the ground terminal of a PC audio out is another). Try this test both with the input shorted and with it open. You might need to turn the volume way up to hear something if the amp is really good.
The audibility of certain effects depend on many factors. Recently I've adjusted input impedance curves. Does it make an audible difference? Not earth shaking, but adds some nice image depth and focus characteristics. The resulting sound is much more seducing. As usual, the mod goes into some other test units to get feedback. I don't believe in golden ears, but I do believe in what ordinary users will hear without giving opinion of my own, then just plainly ask whether they hear any difference or not, if they do, then just ask them to describe what they hear. Really don't know what this does to distortion figures yet, and the impedance is not even matched by strict measurement.
Last edited:
I find it surprising that topology does not correlate with measured sound quality. Am I missing something?
You are sure they have some relation with each other? That the golden ears are describing a physical property of the sound? I know tasting wines is an exact science, but the descriptions of the golden ear types sound much like poetry. You read things into it, so it doesn't need to say anything per se. It's just about what the description makes you feel the sound quality is.
The argument by Lynn Olson was that the valve (specifically the triode) is the most inherently linear active device there is, so with it one can build amplifiers with no feedback at all.
Someone with "golden ears" is defined as a person who pulls this kind of descriptions out of thin air. There is no way your description will mean anything to anyone. There is no way for you to convince anyone that the difference you think you are hearing is anything else but your imagination.
People have heard enormous differences when there have provably been none. Also, whenever I read anything written by you, I feel like you were here for the sake of pushing a religion. I already have one. And one of the greatest things in a good religion is to make a distinction between meaningful and meaningless.
Some people just want to share what they feel and what they experience. They will go on from a friend to a friend, giving out their innermost details. Often that means they are not really giving anything valuable to their friends. Instead, their friends are doing a valuable service by giving a forum (lending an ear) to the person's thoughts and feelings. It's more a "I like to hear me talk, so please allow me" kind of thing.
The correlati0on would be between the specific design and sound quality but not the topology. In other words its possible to make a high quality design out of any topology. Maybe thats what you were missing?
For example class-A topology has an inherate advantage in crossover distortion because it doesn't have any. But that design has so many other problems that are just too hard to get arround.
Class-D is inherantly a disaster because of the complexity of making a broad band highly linear design. But that does not mean that it canoot be done, its just complicated.
Class-AB has many advantages, but crossover distortion is NOT one of them. Not properly handled crossover distortion can kill a class AB amp. However, just like anything else, proper design can make a class-AB amp as good as a class A in terms of sound quality.
The correlation is all with the implimentation and not with the topology.
I agree with you on this, Gedlee. It is almost impossible to make a powerful class A power amp, BUT the transition between class A and class AB is VERY tough.
No, this is science, not religion - always there will be uncertainty.
In that case I wouldn't be able to use it to contribute to my understanding. I think we must have been reading differing accounts from listening sessions🙂 By the way, I'm not sure that tasting wines is an exact science but they have a fairly precise vocabulary to describe aspects of the taste and nose of a wine. This vocabulary is far more well developed and widely accepted than audiophile vocabulary at present. So yes, wine tasting is considerably more exact than audiophile listening - there are no fiendishly difficult 'Masters of Stereo' exams for a start😛
I agree that soongsc's account is not one I could use myself - it is too short on details - like equipment set-up, music used. So it could easily be placebo induced. That's not at all my idea of a 'golden eared' listening account.
I read somewhere once that wine tasters base much of their comments on the appearance of the wine, while sincerely believing that they are just talking about the smell and taste. Even when shown evidence for this they deny it.
Of course they deny it! Just as Golden Ears deny reality when they are faced withn it. In both cases these "experts" have been debunked in blind tests. So, they just claim that blind testing is at fault. You can't win when faced with irrational arguments.
There are different kinds of competitions and procedures for wine judging. However in some there are at least 3 categories that require a visual inspection 1) clarity, 2) color, 3) "legs".
I agree that you are probably right that judgements in these categories are probably not independent from the other taste-kinds of judgements.
- Status
- Not open for further replies.