Purifi Class D

[Jud]

5 hours ago, Iving said:

I don't feel strongly about you saying, "I doubt, based on the scientific work of the past decades, that ABX testing is a very good tool for teasing out any differences in equipment that may still exist."

 

See https://en.wikipedia.org/wiki/New_Coke .  In a blind A/B test, the sweeter drink won.  Yet when a sweeter drink was marketed to consumers, it bombed. Think about why this happened - I believe it has implications for audio.

 

5 hours ago, Iving said:

Probably - there is only one "truth"

 

Left-handers often hear differently than right-handers, see http://deutsch.ucsd.edu/psychology/pages.php?i=208 .  But fascinatingly, some left-handers, apparently as a result of experiencing our right-handed world, hear as right-handers do. I believe this also has implications for audio.

[fas42]

It is remarkably easy to determine whether a system is reproducing the source highly accurately - but most seem not to be aware of what to listen for, or refuse to accept that such can be a valid method, or haven't done some 'training' to make their hearing sensitive to what is going on. Warning: doing this can make you forever aware of the typical distortion that rigs generate, and spoil the listening when experiencing sub-par systems.

 

Very easy first step: select a "bad" or "difficult" recording you have, and wind up the volume to a solid level; then listen a foot or two away from the front of one speaker. If you feel you're being assaulted by a noise weapon, that there is a vicious glare doing its best to damage your eardrums, then congratulations - you are now aware of the "sound of playback distortion". If you don't experience this, and all you hear is clear music, then also you're in a good space: either the system is extremely accurate; or, your human hearing system is wired differently, and so all of the above doesn't apply.

[Iving]

12 hours ago, Jud said:

See https://en.wikipedia.org/wiki/New_Coke .  In a blind A/B test, the sweeter drink won.  Yet when a sweeter drink was marketed to consumers, it bombed. Think about why this happened - I believe it has implications for audio.

 

Well - evidently people buy stuff for reasons other than what they may "objectively" prefer [many drinkers reacted angrily to finding they had chosen a brand other than their favorite]. I say "objectively" in inverted commas because a (lab) blind taste test may demonstrate that "all things equal" A is preferred over B or C - but, evidently, things are not otherwise equal. In the case of cola drinks people are swayed by brand - i.e. what they understand. Possible in Hi-Fi too I guess. "I'm a dCS guy." "Can't hear the music unless it's piped MQA." What are the factors that promote brand over "objective" preference? People may want to "belong" (In Crowd) or feel superior (I only drink Coke.) They may simply seek nostalgia - and why not - just love the shape of that Coke bottle - I am right back on the beach listening to Mungo Jerry at Chart position #1. Not all people of course! There are individual differences - shades of effect - iow "Personality". Main dimension something to do with suggestibility. Now suggestibility doesn't necessarily mean that Subjective experiences aren't "real". Of course they are! If someone buys a $/£/€10k cable and believes they're getting their money's worth, that person may be a suggestible oaf - but they have - indeed - gotten value for their money. Their Subjective experience is no less real than yours or my preference for Buddy Holly or Britney Spears. Of greater interest is whether leverage can be obtained on those Subjective experiences - they are rendered elastic/plastic - such as to conform with what "objective" data determine we should prefer. That is the mission of ASR. And it is a mission. The ASR onus of proof is always on the golden-eared audiophool whom ASR people are "desperate" to convert. People polarise in order to be right. It is the cause of all human conflict - the reverse of "Live And Let Live". "Women are better than men." "Men are better than women." OK a dated example - but for illustration. "ABX tests are the only source of listening objectivity." "Ultimately I trust my ears." People cleaving to these polar positions possess a reality of sorts - if they truly believe what they say; however, by all accounts progression leads us to understand that men and women are merely different regards biological sex. Actually morally equal? I say with question mark because dinosaurs exist. Many things can be true at once. But polar positions cannot co-exist as "truth" at once. This is what I mean by, Probably - there is only one "truth". If Subjectivists and Objectivists (and people in between) are sufficiently open-minded, that "truth" can be sought co-operatively. It can be hunted iteratively. The Scientific Method - as we know it - may not take us the whole way. But it is a start - and it cannot be outranked by ignorance or prejudice en route. Very often we can get nearer to a better understanding of something by ruling out a lot of ideas we have once taken for granted.

 

12 hours ago, Jud said:

Left-handers often hear differently than right-handers, see http://deutsch.ucsd.edu/psychology/pages.php?i=208 .  But fascinatingly, some left-handers, apparently as a result of experiencing our right-handed world, hear as right-handers do. I believe this also has implications for audio.

 

The Cambiata illusion is fascinating - but one of many examples of brain at work. There is no-one on the planet who understands (to any degree of completeness) how the human brain works. It is full of mysteries. Science would have nothing to discover without mysteries. I have already said that Science, as we know it, may not go all the way in facilitating humans to understanding "everything" - and even that notion - that humans have the capacity to understand "everything" (including themselves) is highly moot as far as I am concerned. Science as religion, aka Scientism, is just as phoolish as the $/£/€10k cable purchase. So I think the greatest asset we can develop as audiophiles interested in these things is open-mindedness. The second prerogative is social nicety aka "courtesy"; at least - the motivation to find common ground instead of seeking to be "right" which - it seems to me - is no different to drinking Coke when you "actually" prefer something sweeter.

 

You asked me to think about your links, which I have done. Twice you said, "I believe ... has implications for audio." Please don't be reticent. What are the implications as you see them?

[Jud]

2 hours ago, Iving said:

What are the implications as you see them?

New Coke: Two implications.

One: The qualities that people prefer in a blind test, a sort of “quick hit” to our sensory perceptions, may not be those they prefer in situations where they are able to partake of the product at their leisure. Thus we have greater sweetness that wins a quick taste test, but not something that consumers preferred over the longer term. In audio, we know that greater loudness will be preferred; if loudness is carefully equalized, are there other qualities that are preferred in rapid listening tests, but not over the longer term?

 

Two: As I’ve mentioned before, humans are terrific at pattern matching. This is why training works in audio as well as other areas. With New Coke, though greater sweetness was preferred in the unfamiliar environment of a taste test, when having a soft drink in familiar situations, New Coke didn’t match the pattern built up over years of “having a Coke.” It didn’t taste “right.” The same with audio. We build up preferences over years and then recognize when something doesn’t fit the familiar pattern. We can perceive this as an improvement, or perhaps as sounding worse. But we should recognize that long familiarity does ingrain patterns that sound “right” to us.

 

Cambiata: 

 

One: Our brains are so good at matching patterns they sometimes create patterns that don’t exist. Thus illusions. (The stereo effect is one.)

 

Two: Patterns created by long familiarity are so powerful they can overcome even our physical neurology, for example some left-handers hearing the Cambiata illusion in the way right-handers do.

[PYP]

7 hours ago, Iving said:

So I think the greatest asset we can develop as audiophiles interested in these things is open-mindedness. The second prerogative is social nicety aka "courtesy"; at least - the motivation to find common ground instead of seeking to be "right" which - it seems to me - is no different to drinking Coke when you "actually" prefer something sweeter.

could substitute "humans" for "audiophiles" and broaden your excellent advice.  There seems to be a sudden fondness on our little planet for hardening opinions as a show of strength when it in fact demonstrates fear.   

[Iving]

5 hours ago, Jud said:

The qualities that people prefer in a blind test, a sort of “quick hit” to our sensory perceptions, may not be those they prefer in situations where they are able to partake of the product at their leisure.

 

Yes - I can think of reasons why the lab cannot recreate conditions that could aid discrimination at home. Context is a big deal in associative learning.

 

5 hours ago, Jud said:

In audio, we know that greater loudness will be preferred; if loudness is carefully equalized, are there other qualities that are preferred in rapid listening tests, but not over the longer term?

 

Well ABX gurus insist (correctly) on equalising loudness across conditions. Your remark suggests that people may succeed in the lab because of factors that do not apply to discrimination at home - the converse of above.

 

5 hours ago, Jud said:

humans are terrific at pattern matching ... New Coke didn’t match the pattern built up over years of “having a Coke.” It didn’t taste “right.” The same with audio. We build up preferences over years and then recognize when something doesn’t fit the familiar pattern. We can perceive this as an improvement, or perhaps as sounding worse. But we should recognize that long familiarity does ingrain patterns that sound “right” to us.

 

Totally see this. Alludes to capacity of brain to perform all sorts of somersaults. Question is whether ABX etc insensitive to pattern matching wrt being able to "tell the difference". Is it possible to develop "pattern-matching" tests which reach the parts ABX cannot.

 

5 hours ago, Jud said:

Our brains are so good at matching patterns they sometimes create patterns that don’t exist.

 

For sure. Our brains make up all kinds of sh*t. Perhaps in some ways we enjoy music more because of it. Another reason for the IBS crowd to leave everyone else be. What might brain inventiveness tell us about "audio reality" either way. Does it matter (to audiophiles).

[Iving]

18 minutes ago, PYP said:

could substitute "humans" for "audiophiles" and broaden your excellent advice.  There seems to be a sudden fondness on our little planet for hardening opinions as a show of strength when it in fact demonstrates fear.   

 

So agree. :-)

[fas42]

4 hours ago, Iving said:

 

 

       Our brains are so good at matching patterns they sometimes create patterns that don’t exist.

 

 

For sure. Our brains make up all kinds of sh*t. Perhaps in some ways we enjoy music more because of it. Another reason for the IBS crowd to leave everyone else be. What might brain inventiveness tell us about "audio reality" either way. Does it matter (to audiophiles).

 

"Making up all kinds of sh*t" is exactly why human hearing is so adept at understanding what's going on in a sound field ... an initial book was written on this, Auditory Scene Analysis: The Perceptual Organization of Sound, by Bregman, which has kicked off a very active field of research. Basically, the brain "fools itself" all the time; meaning we only need the bare details to fill in a full picture; only if something is out of place, doesn't fit the pattern, are we unsettled - and then we reject the 'flawed' clues ... sound familiar, ? This has advanced to the point of creating sound fields that exploit this; deliberately conjuring up an auditory sense of something, which if you analyse the actual sound in the room doesn't exist at all ...

 

[PYP]

11 hours ago, Jud said:

New Coke: Two implications.

One: The qualities that people prefer in a blind test, a sort of “quick hit” to our sensory perceptions, may not be those they prefer in situations where they are able to partake of the product at their leisure. Thus we have greater sweetness that wins a quick taste test, but not something that consumers preferred over the longer term. In audio, we know that greater loudness will be preferred; if loudness is carefully equalized, are there other qualities that are preferred in rapid listening tests, but not over the longer term?

 

Two: As I’ve mentioned before, humans are terrific at pattern matching. This is why training works in audio as well as other areas. With New Coke, though greater sweetness was preferred in the unfamiliar environment of a taste test, when having a soft drink in familiar situations, New Coke didn’t match the pattern built up over years of “having a Coke.” It didn’t taste “right.” The same with audio. We build up preferences over years and then recognize when something doesn’t fit the familiar pattern. We can perceive this as an improvement, or perhaps as sounding worse. But we should recognize that long familiarity does ingrain patterns that sound “right” to us.

Sounds right to me but my brain intruded as I read the New Coke story and came up with the following thoughts/questions:

 

-- Did human evolution stop in the mid 80s?

-- The (U.S.) Civil War will never truly end.

-- For some folks, the War of Northern Aggression is related to Communists.

-- New York and Los Angeles are not to be trusted.  Ever.  

-- Bill Cosby has some nerve!

-- Do marketing guys ever get it right?  Or do they just take credit for circumstantial success and blame somebody else for failures?  

-- Why do some executives make loads of money even though they lack common sense?

[DuckToller]

On 7/4/2023 at 2:15 AM, PYP said:

Sounds right to me but my brain intruded as I read the New Coke story and came up with the following thoughts/questions:

 

-- Did human evolution stop in the mid 80s?

-- The (U.S.) Civil War will never truly end.

-- For some folks, the War of Northern Aggression is related to Communists.

-- New York and Los Angeles are not to be trusted.  Ever.  

-- Bill Cosby has some nerve!

-- Do marketing guys ever get it right?  Or do they just take credit for circumstantial success and blame somebody else for failures?  

-- Why do some executives make loads of money even though they lack common sense?

Please add: Where is Bielefeld ? ( a common German misconception ...)

[PYP]

41 minutes ago, DuckToller said:

Please add: Where is Bielefeld ? ( a common German misconception ...)

The fact that Bielefeld did not appear in the New Coke article if further proof it doesn't exist.  

[DuckToller]

40 minutes ago, PYP said:

The fact that Bielefeld did not appear in the New Coke article if further proof it doesn't exist.  

Absolutely 😉!!!

[Allan F]

On 7/3/2023 at 5:15 PM, PYP said:

With New Coke, though greater sweetness was preferred in the unfamiliar environment of a taste test, when having a soft drink in familiar situations, New Coke didn’t match the pattern built up over years of “having a Coke.” It didn’t taste “right.”

 

IMO, and that of many others I know, New Coke sucked from the first taste test in any environment. It didn't need pattern matching to discover that it didn't taste "right". 🙂
 

[audiobomber]

I actually took the Pepsi Challenge in college. I was a dedicated Coke drinker and chose Pepsi in the blind test. Naturally I assumed I must like Pepsi better, so I switched. Not for long, because I actually do have a strong preference for Coke (i.e. Coke Classic). If I can't get Coke, I'll choose some other soft drink, not Pepsi.  

 

Which jibes with my mistrust of audio ABX tests. They are unnatural and unreliable.

[Jud]

2 hours ago, Allan F said:

 

IMO, and that of many others I know, New Coke sucked from the first taste test in any environment. It didn't need pattern matching to discover that it didn't taste "right". 🙂
 

Seemed very obvious when you tried one and it was disappointingly unlike all those Cokes you’d had your whole life. But thousands of Coke drinkers all across the country chose the sweeter Pepsi in a blind test, a short-term comparison, and that was the genesis of New Coke.

 

Lesson? A/B blind comparisons use fundamentally different mental capabilities than determining whether some sensory stimulus fits a long-familiar pattern.

[audiobomber]

I heard a Coke executive interviewed just after this debacle. He said they added a tablespoon of sugar and more bubbles to make the Pepsi taste-alike called New Coke.

[Allan F]

5 hours ago, Jud said:

But thousands of Coke drinkers all across the country chose the sweeter Pepsi in a blind test, a short-term comparison, and that was the genesis of New Coke.

 

FWIW I was one of those who took the "blind" Pepsi Challenge", but I immediately and correctly identified the Coke, which I preferred. I declined the free Pepsi I was offered as a "reward". 🙂

[FIndingit]

When asked about the flavour of Coke or Pepsi, people have a difficulty describing that's it's little more than citrus, vanilla and cinnamon with sweetness.

[Allan F]

On 7/12/2023 at 1:23 PM, FIndingit said:

When asked about the flavour of Coke or Pepsi, people have a difficulty describing that's it's little more than citrus, vanilla and cinnamon with sweetness.

 

So what! I suspect that many, being aware that the formula is a highly guarded trade secret, wouldn't really care. They might simply say that each tastes like a cola - which it is. 🙂

[audiobomber]

On 6/18/2023 at 5:36 PM, mocenigo said:

Sadly, a course in Fourier analysis and signal processing is a bit too complicated to fit in a message on this board.

And a bit simplistic to describe sound quality:

 

 

[Jud]

1 hour ago, audiobomber said:

And a bit simplistic to describe sound quality:

 

 

If read carefully, the article doesn’t actually say Fourier analysis is insufficient, so the title is a bit misleading. What it says is that adding adequate harmonics may be difficult in practice. I would need to do a lot more research into this lab’s work before I would be ready to accept that conclusion.

[mocenigo]

2 hours ago, Jud said:

If read carefully, the article doesn’t actually say Fourier analysis is insufficient, so the title is a bit misleading. What it says is that adding adequate harmonics may be difficult in practice. I would need to do a lot more research into this lab’s work before I would be ready to accept that conclusion.

Thank you: this is what I wanted to write but I was doing errands and got the notification on the phone. 
 

Fact is, Fourier analysis is sufficient and, to a certain approximation, the effect on sinusoids gives the effect on the whole signal - and there are measurements that determine how good this approximation is. And these measurements are performed.
 

The “new” representation may be simply a more compact one. Which is important, for instance, in order to provide audio compression with less artifacts. This is the very first thing I thought while reading that article. 

[Jud]

Part of the reason for my skepticism is that virtually all recording is done using sigma-delta modulation these days, meaning even commercial productions start at a sample rate of at least 2.8mHz. That’s more than 100x 20kHz, which should leave room for a whole heck of a lot of harmonics - far more than your speakers could hope to reproduce.

 

Of course what happens along the way to the consumer affects this, but that doesn’t mean the basic mathematics of Fourier analysis is incorrect or inadequate.

[audiobomber]

5 hours ago, Jud said:

If read carefully, the article doesn’t actually say Fourier analysis is insufficient, so the title is a bit misleading. What it says is that adding adequate harmonics may be difficult in practice. I would need to do a lot more research into this lab’s work before I would be ready to accept that conclusion.

I disagree. Fourier only contains sine waves, does not account for transient performance, which has only introduced in this century.

 

One solution is to divide sound into two types of components, sines and noise, with a smaller number of whistling sine waves and combined with variable noises, or hisses, to complete the imitation.

 A third component, named transient, was introduced around the year 2000 to help model the sharpness of such sounds. 

 Transients alone sound like clicks. From then on, sound has been often divided into three components: sines, noise, and transients.

 

[Jud]

12 minutes ago, audiobomber said:

I disagree. Fourier only contains sine waves, does not account for transient performance, which has only introduced in this century.

 

You are certainly free to disagree, but that's just mathematically incorrect. It's like disagreeing that 2+2=4, though on a more complex level.

 

What the article is talking about is not a new or unique problem, but rather an issue that's been a popular topic in digital audio circles for decades: How best to reproduce the "inharmonic"/transient portions of sounds. "Inharmonics" are somewhat misleadingly named, because they are in fact made up of harmonics, but odd order ones, often higher odd order ones.  That's just a fancy way of saying that along with whatever the fundamental frequency is, the sound includes 7x, 9x, 11x, even 13x that fundamental frequency. But these odd order harmonic frequencies themselves are each sine waves.  It's just that when they all come together and interfere with each other they don't look very sine wave-ish.  These "inharmonics" dominate percussion sounds, and form the initial attack portions of the sounds of other instruments, like the pluck of a string, the hammers hitting piano strings, consonant sounds in vocals, etc. If the inharmonics are brief enough, like a string pluck, we call them transients. But everything is built up from sine waves.