Digital Interconnects

I think that's your challenge, not mine.

Prove that termination issues caused by cables cause jitter. Prove that that jitter causes distortion of the analogue output. Ideally prove that that distortion is audible, but given we can measure the distortion and know how to avoid it, whether we can actually hear it or not is of secondary importance.

Noise obviously causes jitter in the recovered clock. Does that jitter appear as an artifact of the analogue output?

FWIW on neither of my two CD players (rather old I grant you) is the S/PDIF output connected to any of the chassis, mains earth or analogue signal ground.

Paul

 

Hi,

You will find that in other applications it does and the laws of physics don't just take a quick holiday because audio is involved.

Evidence for that is pretty widely published, from HFN & Stereophile to JAES.

Many that I have come across over time do. Many that show on the schematic no connection use neverthelss the kind of RCA connector where the shell also connects to the screw which holds the thing to the Chassis, thus adding the link.

Depending on age items may predate current CE, FTC and EMI regulations.

Ciao T

 

This is absolute Bo**ocks

I have the beginnings of high-frequency hearing loss. The last thing I want is an enhanced high-frequency prosthesis from an audio system because:
- it doesn't sound like real life, real music, real instruments, real musicians (this is the prime reason)
- it lowers the dynamic range available between (what you would believe I'd percieve as) 'normal' listening levels and the point at which my ear/brain system really starts distorting

How dare you tell me what I 'will tend to prefer'...

When I was younger (not much younger) I could readily hear:
- 19kHz pilot tones on FM radios
- tube whistle from TV sets/CRT monitors
- whistle (pain) from bird/animal scarers
Sadly this is no longer the case.

But my hearing still tells me that I still struggle to find CD replay systems that match what I hear in real life. And where do I hear the greatest disparity? In the extreme treble (sibilance, cymbals). And what do I hear from speakers with enhanced treble? Even more distorted unnatural treble.

I have never ever seen any information or experienced anything from my observation of other people's likes or dislikes in audio systems to substantiate this claim.

Sadly, I feel this demonstrates just how little you understand about what people really hear.

I also suspect, and this is worrying, that if you did hear any difference between, say, digital interconnect cables, then you would immediately convince yourself that you hadn't. Surreal.

Off Topic - Subjectivism and repeatably detectable differences...
Let's go to B&Q and mix a couple of similar colours using the Dulux custom colours thingy. We will mix a pair of colours that are only just distinguishable when directly comparing the two side by side. We will demonstrate that these are distinguishable by getting individuals to view the two colours side by side and demonstrate that they can correctly and repeatably identify them as being different in a statistically significant manner.

Let's now take these two colours and separate them, paint additional swatches with them, shuffle them and randomly mix the swatches. Let's get a random set of people and ask them to do serial double-blind tests - ie. only show them the swatches one after the other. This is important, because this is the only way we can do audio comparisons - one after the other.

Now show me the results of the second set of tests. Show me that subjects can readily identify the differences between the two colours with anything approaching the statistical significance expected of db audio tests.

Let's confuse the issue still further. Let's repeat the experiments using blues, reds, yellows, darker colours, lighter colours, metallics gloss and matt colours.

When we created the two colours we deliberately and clearly statistically demonstrated that particular individuals could determine repeatable difference between the two colours. They are subjectively different. And statistically they are measureably different

Sometimes when particular individuals tell us they can hear differences between particular items they really can. Sometimes when measurements tell us there are 'no' differences, it may be that we haven't yet worked out what measurements to take. I was told CDs were audibly perfect when they first came out I was told that I could only detect volume level differences of several dB. I'm so pleased that our technical understanding of audibility has progressed somewhat over the last few decades.

Sometimes, the subjective observation is correct and the observation itself is a valid measurement. Sometimes it isn't. When enough people make similar observations, it's worth paying attention and trying to understand more. The difficulty is trying to work out which bits are worthy of further investigation and which aren't. Too many tests seem to be designed to deliberately disprove that there is a difference rather than to really determine if there's a difference and if so, why.

Back to Digital Interconnects
- Jitter (above certain levels) is demonstrably audible
- In extremis, an interconnect cable can be designed which deliberately introduces jitter at an audible level
- so we can create a pair of cables which are audibly different

Thereafter, the question is how close in (measured technical) performance to each other these two cables need to be before we cannot detect an audible difference.

And it may be that there are other characteristics of the cable which (maybe when interacting with the equipment at either end) affect the audible performance.

Another valid approach is to simply accept that designing the cable to minimise these measurable effects is the sensible way of also ensuring it has no audible effects. That seems (to me) to be the approach that Thorsten is suggesting here. And if the design proposed achieves this in a cost effective manner (which is after all the goal of good design, or have I missed something) then why decry it or quibble over the calculations and design concepts involved?


I'm going away for a few days now, too much of my life has been spent writing this post, and I'm extremely busy at work, so - have a chat amongst yourselves.

 

I think that are evidences for this and when you come back from sulky you could read page 7-8 of this Dr. Floyd E. Toole article. References for the actual studies are also provided at the end of the article if you are really interested.

 

I found this photo, which has been posted on this forum before,



It shows the S/PDIF output from three sources, via a 75 Ohm cable into a 75 Ohm termination then directly into a rather incapable 20MHz oscilloscope. Top is my old Linn Karik, middle a MidiMan Flying Calf ADC, bottom an early 90s Arcam Alpha. I suppose I should have tried it unterminated, maybe next week when I've some time.

Anyway I think that compensating for badly engineered transports and DACS with expensive cabling is a fools game. Thorsten's cheap 75 Ohm cable recommendation is clearly a good one. Although if 10cm doesn't work 10m might.

Paul

 

Hi,

I would think that non of these are too good. There also seems to be some noise over them, especially bottom trace.

Try an RCA plug & socket (make sure to use the "high end" plugs and socets with a lot of metal in them) at the end of the cable and then a few inch of non CI-specified cable to the 75 Ohm resistor and then the 'scope. You have then an average DAC.

I do not think that you CAN compensate for anything going wrong in cabeling, you can merely attempt to "engineer out" as many imperfections which you consider as being able to cause problems.

The concept of expensive is irelevant in the context, it is one for the market.

HOWEVER, as long as people contine to hear real differences and as long as the engineering establishment refuses to take them serious and dissmisses their experiences as auditory halucinations they will contine to buy what they feel works for them.

The lack of understanding what happens in terms of undesirable, parasitary and other interactions in the Engineering community and the unwillingness to remedy this lack and to act on the results is the single greatest reason "High End Cable" merchants contine to charge huge sums (note, 3DS does make cables) and get away with it.

They fill a clear demand in the market and the usual lack of industrial methodes in production (which would help massively to drive prices down) means you buy "boutique" items at boutique prices.

I'd much prefer to tell our own customers to go and buy inexpensive, off the shelf cables (or even expensive ones from the competition), but nothing I have had in hand lives up to the standards I require, so I end up making them as well (BTW, the basic recipies for all my cables are in the public domain, placed there by me, so anyone can build my cables, if only they would).

Ciao T

 

Handheld digital camera onto a 20Mhz scope with necessarily a relatively long exposure. They're as good as can be expected, I don't know whether the ringing on the top trace is real or an artifact of the scope. Anyway it's the clear differences that I wanted to point out.

Paul

 

So much for measurements in hi-fi.

 

A really good analogy, same in my field - displays - really difficult to spot differences in a sequential A-B, very easy in a side by side eg. you can see <1% difference in luminance error between two blocks of a display (mismatched column drivers in LCD) but a gradual change a cross the display can still be undetected at >30% (typical CRT).

Cheers

Jason

 

"Introduction to critical thinking 101"

If you say "this is like that, therefore <claims>", you need to prove this is like that before your claims have any validity.

This means that you have the collosal problem of showing that the human auditory system is like the human visual system. Until you can demonstrate that, anything else you say founded on your analogy is baseless.

 

This is simply a "tin foil hat conspiracy theory".

The hearing "real differences" has not been demonstrated in controlled tests, and as such it is fair to label the experiences as "auditory hallucinations". The "golden ears" need to stop carping about how unfair blind testing is and start producing the goods to demonstrate they can hear what they claim.

Provide the evidence of a problem. The cause will be determined and eliminated. It's really that simple.

Given that your berating a whole community, you should give some serious thought to the idea that you might be clueless, and that the lack of understanding is yours not theirs.

I'd also suggest you interview at Analog Devices, Burr Brown/TI, or any major R&D lab, and you'll get some idea of just how high the bar is in engineering. Moreover, these people have a sense of humour and are in need of amusement, and from the way you've presented your case here, the laughing would continue long after the interview had ended.

 

Hi,

You are welcome to any opinion you wish to hold.

However, VIRTUALLY NO controlled listening tests exist that have sufficiently large datasets to allow ANY conclusion to be drawn with reasonable probability, as to what is audible and what is not. The few that exist are outside traditional audio and mainly in the perceptual coding research.

In absence of sufficient data we must conclude that neither the presence nor the absence of difference has been demonstrated (let's not have the same argument about statistics, type A and B errors etc again, it has been done too many times).

Enough people have demonstrated to themselves to a degree sufficient for their personal use that they hear a difference.

This has nothing to do with fair or not.

Statistical testing (and that is what we are talking about here) follows certain rules. Unless for example Audio ABX testers take care to equalise the risk of type A and type B statistical errors and FIRST expose, calibrate and VALIDATE their procedures and setup against known audible differences I shall treat their statistics the way I treat most: "I do not trust statistics that I have not faked myself!".

I'll leave it at that, but the ABX mafia has not even a toe to stand on, statistcally speaking and endlessly whining on about null results that where near certainties anyway, due to simple statistics does not add anything new.

Is your definition of problem "Must be illustrated by being audible in a DB Listening test!"?

If so I'm afraid I lack the funds (and will) to bother organising test that provide sample sizes to allow conclusions to be drawn with a reasonable certainty and life's too short for w@nking about with stuff that does not amount to anything.

The electrical problems are easily illustrated, BTW and are well known and understood if high resolution analogue measurements (for example) are the subject.

It just seems most engineers exect physics to go on holiday just because "it's audio" and "none hears any differences anyway".

Funny, in ANY serious field of electronics outside Audio most of these issues are readily understood and dealt with. I will have to admit that I omited the qualification "Audio Electronics Enginnering community". But I guess it's easier to just ignore issues that would cause major headaches to solve and if they do come up after all insist those who percieve them must be hallucinating.

Not interested. I got my EE degree a long time ago and have worked for a significant (also in audio). Quite frankly, I prefer a 9-5 in Acounting Software Implementation/Support for the rent and to moonlight in High End audio for "beer & kebab" money. For one, it pays a lot more than EE jobs do... ;-P

Something you obviously miss.

Do you REALLY think I am unaware of the orthodox positions and views. And do you really think that if I was interviewing for a job I'd would put forward contentious issues?

It may have escaped you, I am trying to keep things strictly on popular science level (in my third language too at that) for the benefit of the majority of readers here, who do not have EE degrees.

And I cannot fail to note that you have not made a single line of argument or reasoning that actually illustrates where my explanations (given the audience aimed at) where wrong or inaccurate, instead you hide behind the concept "It's inaudible anyway and it's inaudible because small scale listening tests with no statistical power and poor statistical evaluation of produced data did not show audibility!".

Commonly in modern philosophy we call this Sophistry (which is really doing a great disservice to good old greek Sophists), if you ask me....

L8er T

 

OK, I'll play along. Let's suppose that for one reason or another random testing doesn't find enough Golden Ears, and that the Golden Ears do have a case, then all they need to so is "step forward" and be tested - something that they steadfastly decline to do.

Speaking of "perceptual coding", well there's an example of just how much distortion a signal can undergo before it producdes a just noticable audible difference. You might want to compare and contrast the distortion any cable might induce through jitter etc. and the sort of chicanary that happens in <insert name of favourite perceptual encoder>.

When you already know the outcome (by sight) it's easy to predict what it should be (by hearing)..

Propose a better method if you have one.

If they are so "well known" then how do you explain simultaneous "lack of understanding" in the engineering community??

It depends on where you draw the line though - if you are going to include all those little "cottage industry" brigade in "Audio Electronics Enginnering community", then I might agree with you.

OTOH, even suggesting Analog Devices or TI don't know what they are doing is plainly laughable...

I've been busy pointing out that you have a huge gulf to bridge between jitter and it's audible effect. You've been busy putting a "microscope" on cables so to speak, but are failing to argue the bigger picture. You appear to be hoping that showing "there's jitter on them there cables" is going to be enough to proclaim "and therefore it must be audible".

There's no point in fiddling around with transmission line theory, when your argument is easily going to "come of the rails" elsewhere.

I'd probably put it like this: "We know from large scale listening tests what people can and cannot hear. This has data has been reinforced and revalidated time after time in numerous experiements. That data allows us to predict whether 'jitter and it's effects on the DAC output' are audible. It is not. Small scale studies have validated this position by experimenting with extreme levels jitter, where jitter has not been audible upto the point at which the jitter is so high that the receiver losses lock."

Stereophile/HiFi+/HFN all have the means (money) to get 10 people into a room (the Golden Ears) and know a statistical consultant (whose work they pillaged) to put together an experiment which could "wake up" the engineers if it provided a postive result. One experiment of that form, even though you might care to dismiss its scale, if properly conducted, and assuming it provided an audible result, would get serious attention.

All the experiments be they small or large scale - do NOT support your case.

OTOH, The only "evidence" you provide for your case comes from anecdotal evidence from people involved in sighted tests - who really do have no clue about statistical power and think getting it right once is enough...

Have you noticed that everyone that goes to Vegas wins? Why, because the people who win are happy to tell you about it, while the people who "lost their shirt" stay silent. It's the same with cables/cd players etc: the people who think they hear a difference shout about it; whereas the people who don't hear a difference (and got fleeced) keep quiet..

In other words there's a huge statistical bias in your evidence as well...

 

Hi,

Nothing to do with "golden ears".

AND it also illustrated just HOW LITTLE of certain types of distortion applied to a signal prouces a strongly objectionable difference, something that conveniently seems to have escaped your notice.

As with THD, IMD, TIM et al, single number values are poor indication of audibility. And comparing sour peas to lemons in order to suggest they actually not sour, for pears is sophistry, if I want to be charitable.

My problem is not per se blind testing, it is the related to many fundamental flaws in using for audio, with small sample sizes and often individuals with strong opinions on the subject.

Lack of qualification. Issues around noiseloops (one of the key causes for "cable sound" in interconnecting cables at line levels and also material in digital systems) are understood in areas where the enginner views the whole system and must find ways to make it work, instead of insulated pieces of gear (as is common in mainstream audio).

I actually include mainly the "cost engineering" brigade which supply 95% of the stuff out there.

Neither AD, nor BB/TI (nor cirrus or AKM) make complete pieces of equipment, to be used in the context of a domestic audio system with many seperate pieces of mains powered items interconnected with unbalanced connections, they make building blocks (chips) that are incorporated in such devices. The problems of managing conflicts of electrical safety requirements vs. clean signal interconnections are not theirs.

One of the best sources on that subject is Jensen and their (eminently sensible) advise is "use transformer coupled balanced interconnections at pro-audio levels", which is however of limited use if the manufacturer "forgot" to fit such.

Why, because jitter must not be audible? Because inherently flawed (and often IMHO deliberatly or unconciously weighetd against allowing any audibility) and often inaccuratly interpreted tests failed to show such?

Let's be clear, you where so busy insisting that jitter must be inaudible, because that is the dogma and article of faith that of course it's presence or absence does not matter (as in any religion, any evidence to the contrary of that religion is ignored by believers), that even had comitted major gaffes you would not have pointed them out (thusly strengthening you position that you know what goes and I don't)?

NOT BLEEDIN LIKELY NOW, IST IT?

I merely argued that making an unbalanced, S/P-DIF interconnect in a certain manner and fashion will minimise the results of jitter resulting from mistermination inherent to this type of cable and which at the same time will drastically minimise the problems resulting from noise currents circulating between the chassis of interconnected mainspowered equipment.

I also happened to opine that in my experirence such a cable provided better sonics than some not based on the same principles. Of course, I have not done any db tests that have any statistical power as proof, but then again, neither has anyone else (even in terms of negative outturn).

If you where to put it like that then strictly and logically you would seriously overstep the evidence extant.

We know a lot about how likely it is that a given person may or may not hear a certain alteration of the signal, in a certain context and under certain conditions, no more and no less.

One may be able to draw general inferences from some of the data, but most includes datasets that are too small to offer any statistical power and hence are of no consequency in general. Sorry, but that are the rules of the game and you insist on playing it....

I do not present any statistical evidence, so there is no bias. I merely reject any wideranging application of statistics whith practically zip statistical power (in other words cases where the significance level is set inapropriatly resulting in a very high risk of type b statistical erors, often a risk sufficient to reach near certainty of comiting such errors, for a formal definition), while you are happy to accept the same statistics as fact.

I have no wish to change your religeous belif, but the facts do not support them all that well.

Ciao T

 

I'm using "Golden Ears" in the sense of "those who have heard jitter" etc. It seems only prudent to test these people otherwise you would simply claim that there is no purpose in testing people who "cannot hear jitter".

cite a paper/experiment supporting this statement, so that we can all see your working definition of LITTLE, and verify whether or not this result would be predicted by what is known about psycho-acosutics. Or is this observation also based on some idle chit-chat you read on the internet?

I can see why you'd like to believe that because 2% THD and 0.5% IMD are the limits of audibility IIRC from BCJ Moore's book. No doubt he is "wrong" because he is part of the conspiracy.

Anyway, CD players - cheap ones from Denon and Sony - have inconsequential amount of THD. You should know better than to bring up IMD in this context. And TIM has been widely discredited.

No wonder you're not interested in those measurements, they make a mockery of your case..

The results of Blind Testing are not OPINIONS, they can be legitimately labelled as facts.

What are the "fundemental flaws" of blind testing? the only appears to be that in your opinion such testing does not supply the result you wanted.

AD/TI are aware of all the issues in building a product. If you want a different example - Sony.

Oh, that would be Jensen, the manufacturer of transformers? [Sorry, I could help that..]

Yep, it's all one BIG conspiracy. People who could provide the evidence you need have fudged the results and are concealing the proof that stereophile have been right all along.

It's area 51, but in audio.

I could cheerfully replace the inaudible in that sentance with audible and pretty much say the same paragraph to you.

And the argument is unnecessary. You should simply make up such a lead and measure it (or use spice). If the reduction is "drastic" it will show up in the measurements.

Then it will remain an opinion in the "unproven either way" wastebasket of science.

If you want to dismiss what is already known, then the onus is on you to find contradictory evidence. Simply questioning the data (or just stating that it is wrong) isn't going to get you very far, but contradictory evidence might bring you a Nobel Prize.

I'm aware the issue and of Leventhal's work (as abused by Stereophile). What you fail to understand is that if you commit a type 2 error, then the detection probability has to be low. Now from a Science perspective, when the audiophile comes in and promotes "night and day" changes, then the detection probability can be assumed to be high. So, you see the audiophiles are to blame...

Now, if you believe that the detection probabilty is low, you just keep repeating the experiment until either the subject gives up and/or dies, or a statistically significant "positive" result is reached (the audiophile finally heard it..)

Given that people go into hifi stores and lay down several thousand pounds over the period of about an hour based on "clearly audible differences", then we have to assume that the either (a) the detection probabilty is high (b) it's a vanity purchase (c) auditory hallucination.

The facts support my position considerably better than they support your position.

The hearsay, including the various anecdotes that you've read on the internet and in magazines, pretty much contradict all of psychoacoustics as we know it, by the time you eliminate all the contradictions in your position you're left with, well, no fact's at all...

 

Hi,

Geddes/Lee/et al "Auditory Perception of Nonlinear Distortion" is a really good recent one.

Well, in this case BCJ Moore is WRONG because he is wrong.

Try for yourself (blind of course) the sample Geddes/Lee present here:

http://www.gedlee.com/distortion_perception.htm

You see, it is the TYPE (or should we say harmonic profile) of the distortion that matters, not that this is any news....

I am not sure what the "context" is, I note them as items where "single numbers" (eg 0.09% THD, 0.09% IMD etc.) do not tell the whole and have no direct relation to audiblity, something contended since the 50's by people like the BBC's D.E.L. Shorter and N. Crowhurst and illustrated as corrrect by the trials done in the context of perceptual coding.

The RAW DATA produced by blind tests are FACTS. The results of an analysis of this data ARO opinions, based on any number of assumptions and extrapolating from the data. They are NOT facts.

Typhical "Audio" DB Testing is flawed by:

1) Small Sample size
2) Usually poor, to extremely poor actual setup (electronics, acoustics etc.)
3) By usually the individuals partaking KNOWING what is being tested and USUALLY having definitive and strong views on how these things are.
4) Individuals with no interest music reproduction have usually poorely developed analytical facilities where it comes to reproduced sound.

Especially 3) is interesting. I once set up a little DB Tests. All particpants where told we where testing differences between cables. Most participants where "believers", some not so sure, one was a BBC Engineer and absolutely certain that cables made no difference. Instead of testing cables I actually used polarity reversal in one stereo channel. All but our "non believer" scored 10/10 (as they should), our non-bleiver just heard whjat he wanted to hear, no difference and hence scored 5/10. After I revealed what I had done the gentleman was very incenced and has n ot spoken to me hence....

AD & TI may be aware of the issues, but they are not in a position to do anything about it.

As for Sony, please interconnect a Sony Amplifier, a Sony CD Player and a third, earthed source (AV systems are usually earthed theough the aerial). Measure the various groundfault currents and other noise picked in the system.

Of course, you still contend that non of this is audible, not even the 50Hz hum loops, so it would not matter of course.

I agree and that is my point. In absence of reliable and accurate data on the subject which allows conclusions to be drawn with sufficient certainty (of absence of both type A and typa B errors in the conclusion) we have nothing to base our views on other than opinion.

Most likely, Audio has long passed out of the mainstream into the far backwaters and is on both sides of the fence more of a religion with dogma than a science.

THAT is another opinion.

Ciao T

 

So you're saying all CD players sound the same?

Because of the inconsequential levels of thd?

 

Off Topic

If you are saying that my post was written while I was in a sulk, then this comment is inaccurate, inappropriate and unconstructive. Otherwise, I'm sorry if I've misinterpreted your comment, but this really is a use of English language which I don't understand.

This part of the article demonstrates how to perform listening tests to provide useable results when speakers are being developed by a mass market manufacturer. It doesn't demonstrate very much about the listening discrimination of an individual. It says that within a family of loudspeakers, listeners will generally agree as to which is the better. It does not say that "Not surprisingly, people with "good hearing" pretty much like the same thing"

"...most of the time, liked and disliked the same loudspeakers."
not too surprising if measured on a scale of crap to very good. However, wander round a hifi show and see what different individuals like or dislkie, or look at speaker purchases made, or read the discussions on loudspeakers within this forum and tell me that "people with "good hearing" pretty much like the same thing". I still don't see the evidence for this.

To look specifically at this part of the article:

"All of these erratic listeners had hearing loss."
This does not say that all people with hearing loss were erratic.

"Listeners with hearing loss not only exhibit high judgment variability,they can also exhibit strong individualistic biases in their judgments.This comes as no surprise,since such individuals are really in search of a “prosthetic â€Âloudspeaker that somehow compensates for their disability."
This is not a conclusion, this is a hypothesis. No correlation has been demonstrated between the type of hearing loss and the prosthetic compensation of particular equipment. And if true, is once again a generalisation. No attempt has been made to understand why these listeners (hearing loss or not) made the variable ratings. I repeat, I am not looking for compensation for my hearing losses, I am looking to match what I hear in real life.

"If there is any desire to extrapolate the results of a listening evaluation to the population at large,"
No there isn't. The desire is to understand what is audible and what isn't. The paper is looking for:
"an acceptable target audience for most commercial purposes."

"most listeners found it difficult to ... remember how various test products sounded."
Well there's a surprise.

"not all listeners paid attention to all perceptual dimensions,resulting in judgments that were highly selective."
And there's another surprise. But it doesn't make their ability to hear differences, their assessment and selection of preferences wrong.

This section of the paper successfully shows that on average, all things are average... because that's what an average is. It does not show:
"Not surprisingly, people with "good hearing" pretty much like the same thing. "
And it really, really doesn't show:
"People with hearing deficiencies will tend to prefer speakers which act as a prosthesis for their impairment."

 

Off Topic - Subjectivism and repeatably detectable differences...

I make no claims. I have no problems. My analogy is well founded.

I make an observation on the complexity and subtlety of the human analysis machine based on one of its senses. I therefore allow myself to wonder if some of the measurement techniques used to demonstrate and understand the full sensitivity of one of the senses, and the understanding that gives us of the capabilities and limitations of a other measurment techniques, should therefore be used to develop or assess alternative measurement techniques for other senses. And of course other species have senses developed in a way which we have not. Whilst this gives no definite answers I believe it aids my understanding.

Analogies are good. They help our assessment and understanding and allow us to question the world around us and examine it from a new viewpoint. They may not be right.

I believe the key point from my analogy is that the human analysis machine is demonstrably capable of recognising very subtle variations in one dimension/measurement (ie along one colour) that is buried within the single sense that we call vision. And I suspect that those who are expert in the field of vision could tell us of many other dimensions/measurements where perception is similarly subtle. I therefore struggle to believe that:

1. our ability to discriminate between bits of hifi is fully defined and measured by frequency, amplitude and harmonic distortion

2. we have nothing left to learn on the correlation between measurements and what we can hear

So I'm fascinated when there seems to be a genuine attempt to provide new measurements that relate to audibility. Some of these relationships may be valid, some may not. They shouldn't be dismissed out of hand on the basis that 'we know everything already'. In every other field of science, it's plainly obvious that this isn't true.

(I've got a few more analogies up my sleeve, whether you want them or not, when I've enough time to write them down)

 

Hope at least you enjoyed reading the article by E Toole written when working as an advisor for Harmon K. I guess your general assessment is a fair conclusion. I think he was a serious researchers and his AES papers at the end of the article are what I was refering to but being a professional journal they like them to be protected from copyrights and restricted.

By the way, do you (or anyone else) think this forum has become less enjoyable to visit of late? I find people comes here looking for a 'fight' rather then have a chat with friends as it were.