Hi, Janneman,
I also agree with you on this one. 😀 Even a simple thing, the lower harmonics should be canceled or should be leaved alone, you ask 10 top designers (which their product excel in the market), you will get different answers.
I also agree with you on this one. 😀 Even a simple thing, the lower harmonics should be canceled or should be leaved alone, you ask 10 top designers (which their product excel in the market), you will get different answers.
I would say you can completely change the sound characteristics of an amplifier by altering the composition of the first 3-4 harmonics. Much more difference than reducing THD by say 10 times.
Lars Clausen said:
(Lars, written below is not any opposition against you, it's my thoughts about the situation only)
I think, there is 🙂 Take "test cutomers" from different "taste groups" as many as you can, spent month or two or year to perfom thousands of ABX tests with hundreds of AB-pairs. I'm sure, you will discover there isn't so many "taste groups" indeed. There is the only one group. And I'm sure, you will get unambiguous technical ideas about perfect amp. Nothing misticism.
The problem is very simple: If anybody wants to accomplish such massive tests, he hasn't opportunity. If anybody has opportunity, he hasn't a motivation 🙂
anli said:
Anli,
There is also another issue here. People don't buy stuff after ABX db tests. Equipment gets a certain reputation based on a lot of things - I don't think I need to go into THAT. Everybody knows that NP makes the greatest amps in the world, no? Anybody ready to state that, say, Cello amps suck? Or that Wadis DACs are crap? I thought so. And all that without ABX etc!
Jan Didden
janneman said:
Jan,
Probably, such forum is the best place to understand reality rather live inside a MATRIX of marketing 🙂
From a strictly enginnering viewpoint, if one can listen to any harmonic it's not a 'wire with gain'.
It's euphonics, and there will never be an agreement over it.
It's euphonics, and there will never be an agreement over it.
Jorge said:
Indeed. Euphonics sell. Who wants to pay megabucks for a piece of wire?? Euhh, yes, OK, I forgot..
Jan Didden
I agree in theory.
In the real world it's very difficult to make a an ultralow THD amplifier that doesn't sound terrible. High feedback gain is needed to achieve say 0.001% THD under load conditions. This works effectively against the sound quality.
I have only heard one ultralow THD amplifier, that actually had a good and natural sound to it. (Will not say which .. 😎 )
I think that this article represent not science, but scientism, at its best (that is, its worst): the belief that the entire world has already been explained to perfection by current textbook knowledge. The tone strikes me as utterly unscientific (plus, incredibly arrogant), in that I understand science as a method, certainly not a collection of "facts" (ain't no facts in science, only theories temporarily held to be valid approximations to the underlying realities until proven incorrect). I mean, to start with, how do the authors allow themselves to judge what is "perfect" and what problem has been "solved"?
In my experience, in general, the better the scientist / engineer / politician etc, the more careful with his opinions and the less dismissive of dissent he will be. Strong belief in unshakeable truths marks people who aren't quite sure of themselves at all. And, the recently converted.
Now, I completely agree there are many scams in hi end audio, but this is hardly a mass market. The mass market buys iPods and bookshelf stereos with advertised <10% (!) THD and doesn't bother. Or Bose. In other words most people don't care at all.
For those who prefer demonstrably colored amps (or other elements of the chain, especially speakers): I see another alternate theory why that may be a rational choice. After all people listen to music for emotional fulfillment. And maybe, maybe, the average recording really does give the best emotional fulfillment after some alteration. One example: I personally find real violins very sweet and pleasant sounding. At the same time, you hear surprisingly little "detail". But recorded violins rarely sound that sweet. On the other hand you hear all sorts of distracting detail, such as scratching agains fingerboards etc. In close miked cases such as these you will be better of with less detail - less detail actually comes closer to the real performance.
The authors compare audio to visual perception. Well, in photography for once, again many bits of technology (not photographic technique) demonstrable alter the image from the recorded "truth" (haha) . One example: sharpening. Say, you have a digital photo. It looks soft. This because of the Bayer interpolation of the 3 color CCD sensor data into one picture. The software will then "sharpen" the image, by applying an edge enhancement technique. BTW your retina does something similar. The result is not a gain in information (resolution, sharpness), but in perceived sharpness and contrast (acutance). The image is perceived as more detailed, and more "real" this way. So, even in the visual domain, straight "unaltered neutrality" doesn't cut it. Recording and playback are artefacts by definition, and there is no such thing as neutrality for them.
Another example to match the "too much detail" effect I mentioned above in audio: In portrait photography, people use soft lenses. Why?Pportraits look best this way. Microphotography of every pore and pimple does not make a good portrait.
etc etc etc. Basically reality is much deeper, and much more interesting, than the authors of this article seem to believe.
Disclaimer: I actually do seek neutrality and detail, and technical "perfection" in my contraptions. But I also see (sic) that it's not always a blessing for the listening experience, and often more like a game with myself. And I currently use very ordinary chip amps DIYed close to data sheet with garden variety components.
In my experience, in general, the better the scientist / engineer / politician etc, the more careful with his opinions and the less dismissive of dissent he will be. Strong belief in unshakeable truths marks people who aren't quite sure of themselves at all. And, the recently converted.
Now, I completely agree there are many scams in hi end audio, but this is hardly a mass market. The mass market buys iPods and bookshelf stereos with advertised <10% (!) THD and doesn't bother. Or Bose. In other words most people don't care at all.
For those who prefer demonstrably colored amps (or other elements of the chain, especially speakers): I see another alternate theory why that may be a rational choice. After all people listen to music for emotional fulfillment. And maybe, maybe, the average recording really does give the best emotional fulfillment after some alteration. One example: I personally find real violins very sweet and pleasant sounding. At the same time, you hear surprisingly little "detail". But recorded violins rarely sound that sweet. On the other hand you hear all sorts of distracting detail, such as scratching agains fingerboards etc. In close miked cases such as these you will be better of with less detail - less detail actually comes closer to the real performance.
The authors compare audio to visual perception. Well, in photography for once, again many bits of technology (not photographic technique) demonstrable alter the image from the recorded "truth" (haha) . One example: sharpening. Say, you have a digital photo. It looks soft. This because of the Bayer interpolation of the 3 color CCD sensor data into one picture. The software will then "sharpen" the image, by applying an edge enhancement technique. BTW your retina does something similar. The result is not a gain in information (resolution, sharpness), but in perceived sharpness and contrast (acutance). The image is perceived as more detailed, and more "real" this way. So, even in the visual domain, straight "unaltered neutrality" doesn't cut it. Recording and playback are artefacts by definition, and there is no such thing as neutrality for them.
Another example to match the "too much detail" effect I mentioned above in audio: In portrait photography, people use soft lenses. Why?Pportraits look best this way. Microphotography of every pore and pimple does not make a good portrait.
etc etc etc. Basically reality is much deeper, and much more interesting, than the authors of this article seem to believe.
Disclaimer: I actually do seek neutrality and detail, and technical "perfection" in my contraptions. But I also see (sic) that it's not always a blessing for the listening experience, and often more like a game with myself. And I currently use very ordinary chip amps DIYed close to data sheet with garden variety components.
Jan, I didn't try to make amp with " high masking factor ", it's target for commercial firms 😉 and I know several firms with very well known names, which make it. These firms produce amps, which are relative fast and which have high distortion, mainly at high frequencyand this distortion is very depending on output level. This distortion substitute, by listening with digital format, losses of dynamic of this format on high frequency ( all trebles of all instruments sound by these formats the same, listen closely 😱) ). Because this distortion aren't by all levels the same, character of trebles is changed by sustain, similar like in natural " live " sound. These amps are by not so experienced listeners very prefered, 'cos they sound relative natural. All is only about searching of " correct value of this distortion " or about " correct sustaining of this distortion " and these amp you can call as " very musical amp " 😉 Think about it closely and you will see, that I'm correct 😉 .
Mike, you didn't get it:
"The authors compare audio to visual perception. Well, in photography for once, again many bits of technology (not photographic technique) demonstrable alter the image from the recorded "truth" (haha) . One example: sharpening. Say, you have a digital photo. It looks soft. This because of the Bayer interpolation of the 3 color CCD sensor data into one picture. The software will then "sharpen" the image, by applying an edge enhancement technique. BTW your retina does something similar. The result is not a gain in information (resolution, sharpness), but in perceived sharpness and contrast (acutance). The image is perceived as more detailed, and more "real" this way. So, even in the visual domain, straight "unaltered neutrality" doesn't cut it. Recording and playback are artefacts by definition, and there is no such thing as neutrality for them"
They only said: in a visual thing you can point to it and explain why you like this better then that. You can't do that in audio. You completely turn it around, why?
And after you bash them, you continue to agree with them, that taste is all. Maybe you can point out where they are unscientific, where there premises are wrong, why there conclousion is wrong? That way we would all learn something.
Jan Didden
"The authors compare audio to visual perception. Well, in photography for once, again many bits of technology (not photographic technique) demonstrable alter the image from the recorded "truth" (haha) . One example: sharpening. Say, you have a digital photo. It looks soft. This because of the Bayer interpolation of the 3 color CCD sensor data into one picture. The software will then "sharpen" the image, by applying an edge enhancement technique. BTW your retina does something similar. The result is not a gain in information (resolution, sharpness), but in perceived sharpness and contrast (acutance). The image is perceived as more detailed, and more "real" this way. So, even in the visual domain, straight "unaltered neutrality" doesn't cut it. Recording and playback are artefacts by definition, and there is no such thing as neutrality for them"
They only said: in a visual thing you can point to it and explain why you like this better then that. You can't do that in audio. You completely turn it around, why?
And after you bash them, you continue to agree with them, that taste is all. Maybe you can point out where they are unscientific, where there premises are wrong, why there conclousion is wrong? That way we would all learn something.
Jan Didden
OK, there may be situations such games are funny. Let's make perfect amps only (which are sonic neutral). For the games special external devices may (must!) be used 🙂
Hi, Darkfenriz,
Have a drawing?
In earlier post, a member said it is better to have as high as possible VAS impedance, not lower value.
What do you think the merit of having low impedance VAS?
I can imagine it, but I'm not sure I'm right. Is it making a feedback loop to the differential from VAS? (Besides the main NFB loop from the output stage)?
Have a drawing?
In earlier post, a member said it is better to have as high as possible VAS impedance, not lower value.
What do you think the merit of having low impedance VAS?
Jan,
I should point out that
- my name is Markus 😀
- point taken on visual vs audio, I may have implied something that the authors didn't say, and I agree on that differences in visuals are easier to point at (literally). However my main point in my paragraph on photography stands, namely that "perceived closest to reality" is not equal to "factually closest to reality"
Now for what I didn't like about the article. Mainly, the omniscient (by implication, not by statement) and arrogant tone. To me it sounds full of nasty subliminal implications and clearly written with an agenda..
In detail:
"Unfortunately, comparisons of audio equipment at the consumer level are never done with double-blind tests. The 'buff books' that consumers of audio equipment read are also not using controlled subjective listening testing (except The Audio Critic). "
Unwarranted generalization ("never" - huh ????). Used to conclude that "all" subjective reviews "must be wrong" (implication)
"As would be expected, given that the testing methodology is so sloppy, little correlation exists in the descriptions of the sound of a component from reviewer to reviewer."
Data?
"Controlled listening tests have consistently shown that electrical components will be audibly indistinguishable if the have: (1) flat frequency response, (2) noise and distortion levels below audible thresholds, (3) high input impedance and low output impedance [D. Clark 1982]."
Circular argument: indistinguishable if (1)..., (2) *below audible thresholds*=indistinguishable (!), and (3)... That's a funny one.
"So how can the audio magazines justify the differences they report, when the measurements they make on the equipment show that well-designed audio electronics have virtually flat frequency response and noise plus distortion 80 dB or more below the fundamental test signal?"
Well, maybe -80 dB is not enough given the human ear SNR of 120 dB, but no:
"use pseudoscience"
blatant, generalizing ad hominem attack. Nevermind that some measurements may not show differences, other may show a difference, including the ones Mr smart tech didn't perform. Absence of evidence is no evidence for absence.
In the section on "Disinformation for the innocent consumer." I mostly agree on the content, yet again the arrogant presentation puts me off. I don't believe in any of these mysic tweaks either, I believe they are rip offs, I would never buy them, not even test them, but hey, what do I know? Nothing, so I shut up about them. Some may have a strangely unexpected real effect. And what do they know? Nothing either. Did they study effect of these tweaks? No. They dismiss them as "cannot work because, well, cannot work, because you know, if you knew science, you'd know".
On frequency response, well, maybe a lot of listening material *does* benefit from HF rolloff? It never crossed their minds to come to such a much more interesting conclusion. It may not represent an amp-cable-speaker system of technical fidelity, but it may represent a better listening experience. So, in dismissing consumer preferences for high low order harmonic content, and HF rolloff, they conclude that consumers are dunces rather than asking themselves, why oh why do consumers prefer these alterations? The neurosis of the search for the best amp may be best explained by the assumption that a truly perfect reproduction of most recordings would reveal how awful they really are ;-) But such thinking would ask for some peripheral vision which the authors so obviously lack.
The section on "designer's dilemma" is good, or let's say, I like the presentation. Though
"With respect to distortion, most studies have shown that 0.1% distortion is inaudible even when using test tones. Other studies have shown that distortion on the order of 2% is not audible on music signals [D. Clark 1982]."
makes me think of Geddes's proposed metric where he claims that *some* distorition is inaudible at say 3% and *some* other is audible at 0.00x%. Again, generalization from a few studies never helps.
Plenty of snide remarks in what follows.
In the conclusion section:
"The random nature of the designs strongly suggests that no 'X factor' parameter is being optimized. "
Unsound reasoning. The random nature of the designs suggests that either, no X factor exists, or, that many different approaches may optimize the X-factor equally well, or, that none of the various approaches is clearly superior in optimizing it.
"...instead of facing the reality that electronics exhibiting proper measurements are sonically transparent. "
Claim to absolutes, never good in my book.
"The results of the study of the circuits discussed above confirm the results of controlled double-blind tests, which have shown that no sonic differences exist in audio electronics that measure well. "
Data? Citations? I've read diverging reports. Not to mention that proper subjective testing is a can of worms per se..
"When circuits that are claimed to “sound better” are analyzed, random design techniques are noted. "
Which may point to , again, that either, no X factor exists, or, that many different approaches may optimize the X-factor equally well, or, that none of the various approaches is clearly superior in optimizing it.
"No “X facto?’ exists. "
We don't know that, and they haven't proven it.
"Designers are wasting their time developing audio equipment using the “design, listen, design” approach because they are not using controlled techniques in the “listen” part of the process."
How do they know what all these other designers do in their labs? Well, the authors just know.
"If controlled techniques were used, the designers would discover that audio design is no different form other electronic design. It is done with a set of specifications, with paper and pencil, with computer analysis programs, and with laboratory measurements.
"
And cooking is done by shopping for ingredients, applying the recipe to the letter, and serving it. Tasting is a waste of time, since, clearly, the same ingredients thrown in the same pot will always taste the same no matter what random differences in timing of the addition, method of chopping, or minor quality difference in raw materials may exist.
Again, the article is full of partly correct observations, yet the authors in my opinion seem to strangely miss the point all the while they keep a patronizing tone throughout.
I should point out that
- my name is Markus 😀
- point taken on visual vs audio, I may have implied something that the authors didn't say, and I agree on that differences in visuals are easier to point at (literally). However my main point in my paragraph on photography stands, namely that "perceived closest to reality" is not equal to "factually closest to reality"
Now for what I didn't like about the article. Mainly, the omniscient (by implication, not by statement) and arrogant tone. To me it sounds full of nasty subliminal implications and clearly written with an agenda..
In detail:
"Unfortunately, comparisons of audio equipment at the consumer level are never done with double-blind tests. The 'buff books' that consumers of audio equipment read are also not using controlled subjective listening testing (except The Audio Critic). "
Unwarranted generalization ("never" - huh ????). Used to conclude that "all" subjective reviews "must be wrong" (implication)
"As would be expected, given that the testing methodology is so sloppy, little correlation exists in the descriptions of the sound of a component from reviewer to reviewer."
Data?
"Controlled listening tests have consistently shown that electrical components will be audibly indistinguishable if the have: (1) flat frequency response, (2) noise and distortion levels below audible thresholds, (3) high input impedance and low output impedance [D. Clark 1982]."
Circular argument: indistinguishable if (1)..., (2) *below audible thresholds*=indistinguishable (!), and (3)... That's a funny one.
"So how can the audio magazines justify the differences they report, when the measurements they make on the equipment show that well-designed audio electronics have virtually flat frequency response and noise plus distortion 80 dB or more below the fundamental test signal?"
Well, maybe -80 dB is not enough given the human ear SNR of 120 dB, but no:
"use pseudoscience"
blatant, generalizing ad hominem attack. Nevermind that some measurements may not show differences, other may show a difference, including the ones Mr smart tech didn't perform. Absence of evidence is no evidence for absence.
In the section on "Disinformation for the innocent consumer." I mostly agree on the content, yet again the arrogant presentation puts me off. I don't believe in any of these mysic tweaks either, I believe they are rip offs, I would never buy them, not even test them, but hey, what do I know? Nothing, so I shut up about them. Some may have a strangely unexpected real effect. And what do they know? Nothing either. Did they study effect of these tweaks? No. They dismiss them as "cannot work because, well, cannot work, because you know, if you knew science, you'd know".
On frequency response, well, maybe a lot of listening material *does* benefit from HF rolloff? It never crossed their minds to come to such a much more interesting conclusion. It may not represent an amp-cable-speaker system of technical fidelity, but it may represent a better listening experience. So, in dismissing consumer preferences for high low order harmonic content, and HF rolloff, they conclude that consumers are dunces rather than asking themselves, why oh why do consumers prefer these alterations? The neurosis of the search for the best amp may be best explained by the assumption that a truly perfect reproduction of most recordings would reveal how awful they really are ;-) But such thinking would ask for some peripheral vision which the authors so obviously lack.
The section on "designer's dilemma" is good, or let's say, I like the presentation. Though
"With respect to distortion, most studies have shown that 0.1% distortion is inaudible even when using test tones. Other studies have shown that distortion on the order of 2% is not audible on music signals [D. Clark 1982]."
makes me think of Geddes's proposed metric where he claims that *some* distorition is inaudible at say 3% and *some* other is audible at 0.00x%. Again, generalization from a few studies never helps.
Plenty of snide remarks in what follows.
In the conclusion section:
"The random nature of the designs strongly suggests that no 'X factor' parameter is being optimized. "
Unsound reasoning. The random nature of the designs suggests that either, no X factor exists, or, that many different approaches may optimize the X-factor equally well, or, that none of the various approaches is clearly superior in optimizing it.
"...instead of facing the reality that electronics exhibiting proper measurements are sonically transparent. "
Claim to absolutes, never good in my book.
"The results of the study of the circuits discussed above confirm the results of controlled double-blind tests, which have shown that no sonic differences exist in audio electronics that measure well. "
Data? Citations? I've read diverging reports. Not to mention that proper subjective testing is a can of worms per se..
"When circuits that are claimed to “sound better” are analyzed, random design techniques are noted. "
Which may point to , again, that either, no X factor exists, or, that many different approaches may optimize the X-factor equally well, or, that none of the various approaches is clearly superior in optimizing it.
"No “X facto?’ exists. "
We don't know that, and they haven't proven it.
"Designers are wasting their time developing audio equipment using the “design, listen, design” approach because they are not using controlled techniques in the “listen” part of the process."
How do they know what all these other designers do in their labs? Well, the authors just know.
"If controlled techniques were used, the designers would discover that audio design is no different form other electronic design. It is done with a set of specifications, with paper and pencil, with computer analysis programs, and with laboratory measurements.
"
And cooking is done by shopping for ingredients, applying the recipe to the letter, and serving it. Tasting is a waste of time, since, clearly, the same ingredients thrown in the same pot will always taste the same no matter what random differences in timing of the addition, method of chopping, or minor quality difference in raw materials may exist.
Again, the article is full of partly correct observations, yet the authors in my opinion seem to strangely miss the point all the while they keep a patronizing tone throughout.
Actually, the most intriguing observation in the pas few posts was by Upupa - about inherent HF weakness of typical digital format. Upupa, can you elaborate on this? If there's a way to deal with this, it would best be done on the DA level of course, because then you can target the problem in a reproducible way.
Then again, why is it then not done by CDP manyfacturers?
Then again, why is it then not done by CDP manyfacturers?
JORGE
In the same site, according to the author there are the white hats and the black hats :
http://www.biline.ca/critic2.htm
http://www.biline.ca/critic3.htm
Two important persons are missing : Douglas Self and Jean Hiraga.
LARS CLAUSEN
"I would say you can completely change the sound characteristics of an amplifier by altering the composition of the first 3-4 harmonics. Much more difference than reducing THD by say 10 times."
"In the real world it's very difficult to make a an ultralow THD amplifier that doesn't sound terrible. High feedback gain is needed to achieve say 0.001% THD under load conditions. This works effectively against the sound quality. I have only heard one ultralow THD amplifier, that actually had a good and natural sound to it. (Will not say which .. )"
Have you done this experience : add any amount of distorsion using a little circuit with diodes and IC which produces beautifully degraded harmonics (better than a 300B) ahead of an amplifier with distorsion less than 0.001% ? How does it sound ?
~~~~~~~~ Forr
§§§
In the same site, according to the author there are the white hats and the black hats :
http://www.biline.ca/critic2.htm
http://www.biline.ca/critic3.htm
Two important persons are missing : Douglas Self and Jean Hiraga.
LARS CLAUSEN
"I would say you can completely change the sound characteristics of an amplifier by altering the composition of the first 3-4 harmonics. Much more difference than reducing THD by say 10 times."
"In the real world it's very difficult to make a an ultralow THD amplifier that doesn't sound terrible. High feedback gain is needed to achieve say 0.001% THD under load conditions. This works effectively against the sound quality. I have only heard one ultralow THD amplifier, that actually had a good and natural sound to it. (Will not say which .. )"
Have you done this experience : add any amount of distorsion using a little circuit with diodes and IC which produces beautifully degraded harmonics (better than a 300B) ahead of an amplifier with distorsion less than 0.001% ? How does it sound ?
~~~~~~~~ Forr
§§§
- Status
- Not open for further replies.