I know it’s all been said before but this is madness ….

[Jim Audiomisc]

As I understadn your post you means chaotic in the mathematical sense not "chaotic" in the grasping-at-straws-to-justify-anything-goes sense of "producing pretty much any possible output from any given input." -or to put it another way, the circuit may wander round endlessly, but it still does so within a confined space.

The key difference is that real 'chaos' is inherently unpredictable unless you know the starting conditions with *infinite precision*. This can easily apply to a complex analog circuit that satisfies some trivial requirements. 1) Has at least 'three memory values' that affect what comes next. 2) Has an appropiate feeback system so those values affect what comes next.

Chaos more generally tends to mean we don't know what is going on or may happen. However people often use words a la Tweedledee and Twedeldum. 8-]

 

[Jim Audiomisc]

As someone moderately popular in the music world once said “You don’t need a weatherman to know which way the wind blows”.

But you may need one to know if you should buy an umbrella because you'll probably need it tomorrow or Friday.

 

[Jim Audiomisc]

See Audio DiffMaker (libinst.com) for example.

And from way back in time, Baxandall Audible amplifier distortion is not a mystery.pdf | DocDroid

And many other strands. Keith Howard wrote up some experiments and ideas for HFN some years ago. He was locking an ADC to the S/PDIF output of a CD player so aligned captures of different cables could be made with a minimum of complex DSP to diff.

FWIW I don't think there are surprising differences to find, but not finding them is always going to be unconvincing to those inclined to believe. I also don't think it matters given how much interesting stuff there is to explore.

Thanks. Good memory! Yes, IIRC Keith called it something like "Lynx Effect". I did exchange some emails with him at the time, but had forgotten.

In general I'd expect good kit to show very little in the way of problems. But using music as the test signal does help overcome the common objection that sinewaves or impulses might not show up what music gets affected by. We may doubt it is a problem. But test and measure tends to be better than suspect and then assume.

In the end, if someone wants still to refuse to accept results that don't fit what they believe, that's their choice I guess. They can then put their argument and others can in turn decide for themselves

 

[Paul Hindle]

But you may need one to know if you should buy an umbrella because you'll probably need it tomorrow or Friday.

Ah yes but that’s about predictability. New thread?

 

[adamdea]

The key difference is that real 'chaos' is inherently unpredictable unless you know the starting conditions with *infinite precision*. This can easily apply to a complex analog circuit that satisfies some trivial requirements. 1) Has at least 'three memory values' that affect what comes next. 2) Has an appropiate feeback system so those values affect what comes next.

Chaos more generally tends to mean we don't know what is going on or may happen. However people often use words a la Tweedledee and Twedeldum. 8-]

I suppose the point I'm tryign to tease out is whether one can make any prediction about the type 1 chaotic system, or more specfically the analog circuit satisfying those requirements. As you mentioned in a previous post audio devices tend to be modealled as linear plus some higher order effects. As I understand it, those higher effects are limited in their effect though: Linear plus so much noise (or distortion) is not necessarily a problem. We definitely can't know the precise output of the noisy system but we may be able to predict exactly the limits between which its output will fall.

To what extent in practice will the type 1 effects make the circuit behave for pracical purposes differenlty from linear plus defined amount of noise or linear plus so much distortion. Will the effects be unbounded in amplitude or do they creates a defined/definable region of uncertainty?

 

[Joe Hutch]

The key difference is that real 'chaos' is inherently unpredictable unless you know the starting conditions with *infinite precision*. This can easily apply to a complex analog circuit that satisfies some trivial requirements. 1) Has at least 'three memory values' that affect what comes next. 2) Has an appropiate feeback system so those values affect what comes next.

Chaos more generally tends to mean we don't know what is going on or may happen. However people often use words a la Tweedledee and Twedeldum. 8-]

I think you mean Humpty-Dumpty.

But laypeople do tend to use terms like 'chaos theory' and 'quantum mechanics' without any real understanding of what such terms mean outside the specific area of science to which they relate. For example, Mrs H used to work on a psychotherapy journal, and writers would chuck these phrases about to just mean 'something I don't understand/can't explain'. Another favourite was introducing a quotation with 'As Einstein said ...' when there was no evidence that he'd said any such thing.

 

[adamdea]

And nor is this in any way a fair characterisation of my argument. If you're choosing to traduce it rather than engage with it, that just diminishes the perceived validity of your own POV.

Ok I am being grumpy- but where does any of this take us? Will appyign a test to "complex musical signals" produce materially different resuts from existing test using test tones and will this illuminate our understanding of the audibility or othrwise of differences between different cables? (or even power amps)
Yes I think it's interesting to see the delta between the input and output of a system playing music. Yes I think it maybe be useful in illuminating some areas to some people.

But....
There always will be a difference of some amount (whether its thermal noise or whatever). The idea of doing a null test to see whether a device makes an audible difference with a real world signal isn;t entirely new and whenever I've seen it come up it never makes any difference to anyone's position because most if not all of the people requiring reassurance that a real music signal behaves/ does not behave radically differently from a tone or series of tones also refuse to acknwledge that any identified difference can be said not to be audible. And so we go round in circles.

Suppose we take two speaker cables you think sound different and perform the input output comparison of the complex musical signal at the speaker terminals and produce the result that the orgnal and the output null down to -110dB with the resulting difference signal appearig to be noise with maybe the odd tone at the -120 dB level.

What then? Would you conclude that you were probably mistaken or would you start a new "thought experiment"? In what cirumstances would the results of Jim's proposed test make any difference to you?

 

[Filian]

Apart from my still having trouble over what to do with the perfect hi-fi mental experiment, I didn't think I was disagreeing with you on that bit, and definitely not with that post. Mind you, I don't think I have followed the "monumental measurists' self-contradiction" properly yet, but I'll read the whole exchange again carefully and see if that helps.

There is no perfect hifi. One had not therefore been posited. The counter-argument to but-you-can't-hear-that together with its false authorities was offered thus: in a system of sufficient resolution we should allow that everything is potentially audible, or, in a system of sufficient resolution we should first allow that everything is practically audible.

This is a rhetorical argument to restore balance versus the faulty argument that rules what is or is not audible a priori. Personally I've found this the practical case probably ten to one against the opposite, which is to presume and preclude audibility, so I added, because it generally is (audible).

Then, the measurist's self-contradiction is simply that when in the logically faulty audibility argument (declaring X audible and Y not audible) his requirement automatically switches to the rather unscientific demand to produce supporting data to show existence of a phenomenon, the phenomenon allowing audibility. No phenomena, no allowed audibility. Of course phenomena are rather oblivious to this rule and carry on anyway.

Tying the two together, the use of data - or rhetorically speaking per the above, the measurist's rhetorical use of missing data or the data gap - ultimately leads to the expectation that for his presumably inaudible effects, whether via demonstrated phenomena or as-yet unknown phenomena, one should instead select kit with what is presumed to be supporting data. The short version is that we've replaced the ear with the eye and assumed a relevant proxy. We've just moved domains, assuming complete compatibility.

The issues with this are clear, but none as interesting as selecting for audibility what cannot be heard. That's the monumental self-contradiction. Your having consulted a chemist, I'd like you to season my dish, in other words, because I can't taste it anyway and thank you very much for your expertise.

The sheer conviction and expectation freighted in with measurist insistence goes quite out of tune with practical use cases. I'd go one further and say, as I have, that in my experience in real cases, measurist systems have yet to meet, much less exceed, the sounds of setting up the old fashioned way: For systems designed for hearing, simply listening to them. Authentic musicality is, after all, fairly identifiable, which again is the very point.

Nonetheless, this hearing-things-in-audio idea has become a point of some contention.

 

[Jim Audiomisc]

There is no perfect hifi. One had not therefore been posited. The counter-argument to but-you-can't-hear-that together with its false authorities was offered thus: in a system of sufficient resolution we should allow that everything is potentially audible, or, in a system of sufficient resolution we should first allow that everything is practically audible.

Did the initial assertion contain the "potentially" qualifier? Not checked, but I thought the snag was that it didn't. However I do suffer from Liz Dexia to some extent.

For me hi-fi is in practice "perfect" if I can forget its a HiFi and enjoy the music essentially as I please. The main limits for me tends to be things like the room, and unwanted background noise, not the hifi system. Also when cooking dinner or washing up I use a DAP + headphones as that's better than speakers.

Similarly, some may want a hifi that *alters* the source material to be more pleasing than a plain approach. And then fiddle to change this from item to item.

This is why I want things like 'tone controls' to adjust as suit me given variations in source material and what I'd like to hear. Even though reviewers want to NOT have them because they may make mince of their 'wine tasting' when using other kit, source material, and a room quite different to my case.

So 'perfect' is a moveable feast and may require *deliberate* alterations of the source material to approach. Pefect for *what*, for *who*, etc...

Listening to kit *or* measuring it in other situations then becomes problematic as you need more non-kit info to decide what might be judged 'perfect' by the user. This is one reason I regard most 'subjective' reviews as a waste of space. But having my own experience I find some measured info - e.g. noise level or FR - can be useful as a way to help me decide oif an item might suit me.

 

[Sue Pertwee-Tyr]

Ok I am being grumpy- but where does any of this take us? Will appyign a test to "complex musical signals" produce materially different resuts from existing test using test tones and will this illuminate our understanding of the audibility or othrwise of differences between different cables? (or even power amps)
Yes I think it's interesting to see the delta between the input and output of a system playing music. Yes I think it maybe be useful in illuminating some areas to some people.

But....
There always will be a difference of some amount (whether its thermal noise or whatever). The idea of doing a null test to see whether a device makes an audible difference with a real world signal isn;t entirely new and whenever I've seen it come up it never makes any difference to anyone's position because most if not all of the people requiring reassurance that a real music signal behaves/ does not behave radically differently from a tone or series of tones also refuse to acknwledge that any identified difference can be said not to be audible. And so we go round in circles.

Suppose we take two speaker cables you think sound different and perform the input output comparison of the complex musical signal at the speaker terminals and produce the result that the orgnal and the output null down to -110dB with the resulting difference signal appearig to be noise with maybe the odd tone at the -120 dB level.

What then? Would you conclude that you were probably mistaken or would you start a new "thought experiment"? In what cirumstances would the results of Jim's proposed test make any difference to you?

I've never actually seen this sort of analysis performed, so I'm speaking from a position of ignorance (quiet there at the back...) and you seem to be implying that it's been done, and the results aren't particularly remarkable. If it has been done, it'd be interesting to know what the device under test was in those circumstances, and something of the context. If it hasn't been done, or not so much that there's a significant body of evidence, let's not prejudge the findings.

But I have to acknowledge that you're not being unfair in supposing that I'd be reluctant to let go of my viewpoint. I'm not sure anybody would find it easy to deny the experience of their own observations. To which end, I do have a teeny issue with the 'audibility' end of the argument. My preferred way to evaluate equipment is to use it for an extended period, not an A/B comparison. A lot of the things that matter to me are questions of nuance and timing, not gross tonal or bandwidth differences, it's not 'does this sound better/different?' it's 'am I enjoying this more, or less?' and those things sometimes only emerge at a subconscious level, or after protracted exposure. Which leads me to think that on the vexed question of perception, there's a difference between 'audibility' and 'perceptability' and simply stating that 'at -110dB this is inaudible' is an incomplete answer. We know that the brain fills in the gaps, else hifi wouldn't work at all and that suspension of disbelief wouldn't happen so as to bring us music. It feels sometimes like a better hifi lets the brain work less hard at filling in the gaps, though you'd be hard-pressed to isolate any particular element that was missing before, in terms of simple A/B comparisons.

 

[adamdea]

I've never actually seen this sort of analysis performed, so I'm speaking from a position of ignorance (quiet there at the back...) and you seem to be implying that it's been done, and the results aren't particularly remarkable. If it has been done, it'd be interesting to know what the device under test was in those circumstances, and something of the context. If it hasn't been done, or not so much that there's a significant body of evidence, let's not prejudge the findings.

But I have to acknowledge that you're not being unfair in supposing that I'd be reluctant to let go of my viewpoint. I'm not sure anybody would find it easy to deny the experience of their own observations. To which end, I do have a teeny issue with the 'audibility' end of the argument. My preferred way to evaluate equipment is to use it for an extended period, not an A/B comparison. A lot of the things that matter to me are questions of nuance and timing, not gross tonal or bandwidth differences, it's not 'does this sound better/different?' it's 'am I enjoying this more, or less?' and those things sometimes only emerge at a subconscious level, or after protracted exposure. Which leads me to think that on the vexed question of perception, there's a difference between 'audibility' and 'perceptability' and simply stating that 'at -110dB this is inaudible' is an incomplete answer. We know that the brain fills in the gaps, else hifi wouldn't work at all and that suspension of disbelief wouldn't happen so as to bring us music. It feels sometimes like a better hifi lets the brain work less hard at filling in the gaps, though you'd be hard-pressed to isolate any particular element that was missing before, in terms of simple A/B comparisons.

As you will note above- products have existed for some time which compare output and input: audio diffmaker and delta wave etc al.
Try googling. I'm sure you will find some results (certainly a famous Ethan Winer one, but probably others)

But in any event my point stands about the usefulness of the exercise Jim was proposing and which you seemed to be saying you felt was important and connected with your thought experiment. No concept of audibility= no use for anything. In one of his recent posts about this Jim said that if the results showed no difference beyond a level of [x] (which I think was a jokey expresion) , he would not worry about the thing in question.

Now @NickofWimbledon -have we finally got to a practical real life moment to wheel out Karl Popper?

 

[Southeastern]

 

[cooky1257]

Thousands believed Derek Acorah could talk to the dead until the fateful "Mary loves Dick" incident...and I'm of the opinion the more egregious transformative accessory claims merely await theirs...;-)

 

[vesuvian]

I have devised a nice thought experiment for the perpetuators of this thread...

 

[herb]

Similarly, some may want a hifi that *alters* the source material to be more pleasing than a plain approach. And then fiddle to change this from item to item.

Ah flat earth audio, not hifi, which emphasises dynamics which hifi often reduces for the sake of a perfect frequency response or whatever. This is really an old 1970s argument, tone controls smudge the sound as I found out with a FM Radio/Quad/33 303/ESL(57) system all those years ago. I feel old

 

[camverton]

I've never actually seen this sort of analysis performed, so I'm speaking from a position of ignorance (quiet there at the back...) and you seem to be implying that it's been done, and the results aren't particularly remarkable. If it has been done, it'd be interesting to know what the device under test was in those circumstances, and something of the context. If it hasn't been done, or not so much that there's a significant body of evidence, let's not prejudge the findings.

But I have to acknowledge that you're not being unfair in supposing that I'd be reluctant to let go of my viewpoint. I'm not sure anybody would find it easy to deny the experience of their own observations. To which end, I do have a teeny issue with the 'audibility' end of the argument. My preferred way to evaluate equipment is to use it for an extended period, not an A/B comparison. A lot of the things that matter to me are questions of nuance and timing, not gross tonal or bandwidth differences, it's not 'does this sound better/different?' it's 'am I enjoying this more, or less?' and those things sometimes only emerge at a subconscious level, or after protracted exposure. Which leads me to think that on the vexed question of perception, there's a difference between 'audibility' and 'perceptability' and simply stating that 'at -110dB this is inaudible' is an incomplete answer. We know that the brain fills in the gaps, else hifi wouldn't work at all and that suspension of disbelief wouldn't happen so as to bring us music. It feels sometimes like a better hifi lets the brain work less hard at filling in the gaps, though you'd be hard-pressed to isolate any particular element that was missing before, in terms of simple A/B comparisons.

Quite so, and for each individual the system that best suspends disbelief over an extended period of time is likely the best choice for them. Who cares that some online audio pedant can’t measure and quantify that? It doesn’t matter that one persons related experience wouldn’t necessarily apply to any one else, and that doesn’t mean that their related experiences aren’t of interest to others.

Let’s face it, a suite of measurements won’t predict what works best for us as individuals; for one person the latest cardioid speakers might lead to a suspension of disbelief, but they don’t for me. That is the problem with obsessing over measurements to the exclusion of how individuals react to hifi, a hifi system without a listener is somewhat mute. I expect the time will come when we can easily “measure” our hifi in the home by attaching electrodes to our heads and then we can discuss some really worthwhile measurements. Until the I’m going to leave measurements to the designers and use my perception to decide which of their solutions works best for me - apart from using REW to set up my sub of course.

 

[NickofWimbledon]

1. @Filian - I know I have been slow to grasp this, but your 'tying' para helps, so thanks for the patience.

2. @cooky1257 - so you think we are all waiting for Dick?

3. @Jim Audiomisc - I for one am aiming for the most enjoyable music possible. Given my not great hearing and poor concentration and knowledge of the intricacies of a lot of music, and given comparisons with others and my variability, I am a poor test subject. When trying hi-fi over 35 years, the surprise to me is not that I could prefer a less accurate version (or not hear the difference) sometimes - the big surprise is finding how much a little bit more accuracy usually does add to my enjoyment, even when my guessing what is making the difference may be utterly useless.

In any event, neither I nor anyone I know thinks they choose to alter the anything from perceived neutral except to balance something else that is already altering what they hear - usually room or speaker issues. I agree that there are a few who (for example) genuinely prefer a per-Cirkus LP12's 'sound', but are those of us who like neutrality (but are not Objectivists at all) so unusual?

4. @adamdea - I think we are agreeing! Much though I value extended conversations about falsifiability, how to apply Bishop Berkeley and John Locke to the issue of speaker cables, Wittgenstein on the connections between useful language and meaningful concepts and much else, it seems unlikely that this is the place for it. I have friends and whisky for that sort of thing.

5. I suspect that most people here who lean toward a Subjectivist approach will make modest attempts to remove obvious sources of error in testing new kit. The results of all those differently-designed tests on different kit in a different place by different people may be good enough for each of us, though few of us would claim this would contribute to a proof of anything, and most feel no need to go further.

This actually matters if anyone is hoping that 'the industry' in some monolithic sense will fund a really robust test of whether anyone can consistently 'hear a difference' (thus proving that there is one) between a £5pm and a £100pm cables, or between two cables that 'measure the same' but supposedly sound very different, or whatever. Most people buying cables don't need convincing of possible difference (and sometimes the degree to which that is true is disappointing), so which companies could expect how much more in the way of sales if they could show test results that would be significant to the geeks among us? In short, I wouldn't encourage waiting for someone else to do that test.

If a different test that could be done readily might satisfy 1 or 2 Objectivists here (using the term loosely as always), but would be inconclusive to another 5 and not seen by anyone else who was not already a 'believer' in moderately priced cables, who could make a profit out of it? If the answer is still no-one, this may explain why 'the industry' has 'failed' to meet the challenge that some Objectivists think it is their duty to try.

6. I have been trying to think of a way to get from the more adversarial and literally repetitive 'big-picture' discussions here to something that might be useful to anyone else. Thought experiments can be useful when you can't do real ones, though the phrase risks some (as mentioned) name-checking Albert E gratuitously. This is only a first attempt...

If you regard yourself as a strong believer in the difference between evidence and impressions, and sceptical of Woo, imagine that your mum has a decent hi-fi and likes the Carpenters and Debussy a lot - but knows almost nothing of the technology. She mentions that the nice young man at the dealers has suggested she pop in to hear him compare her cheap speaker cables with a new and better product - and, well, the shop is opposite the hairdressers, and £130 doesn't sound much - oh, is that 'per metre'?

We all know in advance that nothing you say is trying to help by deliberately insulting her intelligence in the hope she will suddenly see the light. In fact, you might think or say: "Everyone can spend their money how they choose & don't need my approval". However, you would presumably rate the chances of her spending money for no extra SQ as very high, so you might say rather more.

What would you gently suggest to your mum, as a way of reducing that risk?
Bring the new cables home to try (with a bit of scurrilous swapping when she isn't looking)?
Insist on 3 or 4 tests of old and new at the shop, all keeping her eyes closed?
Or would you be thinking "£200 more or less on cable is small beer and maybe they fit round corners a bit better anyway - I just don't want him showing you an 'amazing' £10K amplifier"?

If cables aren't the place where you think a little common-sense would save the most money, where is that, and what is the least contentious and bothersome way to encourage people to reassess? Is the answer in all cases: "if you must use your ears, and at a dealers, don't be embarrassed - just take a blindfold"?

The reason I ask is because I know that I am a poor tester, subject to all the biases & failings imaginable. Some are better than others, but we all accept we are still fallible, and almost everyone professional is selling something. I therefore think Subjectivists need all the genuine help we can get.

It may well be that we don't all need to agree about 'signal versus noise in cable evaluation' or 'what is sound?' or 'what about chaos?' to be helpful to each other.

If any of us should have had an edge in spotting 'lines-not-to-believe' BS or 'it looks impartial but isn't' tricks, it should be those who lean toward an Objectivist stance. You may or may not think that anyone making judgements by listening is doomed to some extent, but the specific application of your cynicism may still help avoid a few howling errors by others in the future. Wasn't that the idea?

@vesuvian - was your thought experiment a bit simpler, by any chance?

7. It may be harder for Subjectivists to help Objectivists (esp if you are sure we are all a bit potty), but it is not necessarily impossible.

Some want proof (sometimes to their own standard) that there is any phenomenon outside the brain here to discuss. As I understand it, others just wish it was a bit easier to persuade the rest of us to exercise a bit more scepticism and a bit less belief - all for our benefit, not yours.

How about this? Several subjectivists attach extra significance to tests that give a result opposite to what we say we were expecting or that are fairly 'blind', but recording non-results may be suspect and many don't bother trying any of that - if they listen and like, they buy. No-one tabulates the results of tests like that, because they are not 'evidence'. But they could be data...

I am sure we could make a table on a new thread, specifically for 'upgrade' auditions. People could enter what kit is being tested with what, home/ shop, for how long, and what result they expect/ hope for. They can later insert the actual result (in their minds).

That might well have a lot of individually meaningless results, but it would build a dataset. In couple of years, it might well be possible to point to it in some future discussion because (say) '90% of people testing really expensive speaker cables heard what they expected' or 'nearly 30% of people testing speakers at home decide against their expectation' or 'everyone who auditions Shahinian Diapasons (or anything that size) in a shop buys them, but it's only 75% of home auditions'.

This sort of info might give a way for those who can be bothered with these exchanges to point to particularly dodgy/ bias-prone/ Woo-filled issues of which the rest of us should be extra-wary, and to encourage people to be more careful in a way that involves no sneering, no patronising and no mistaken views that either of those was happening. For the more hardcore, it might give data that (for example) comprehensively disproves that results-against-expectation are statistically significant at all.

It's just an idea...

 

[Jim Audiomisc]

Ah flat earth audio, not hifi, which emphasises dynamics which hifi often reduces for the sake of a perfect frequency response or whatever. This is really an old 1970s argument, tone controls smudge the sound as I found out with a FM Radio/Quad/33 303/ESL(57) system all those years ago. I feel old

Depends what you want from a system, and the nature of the input.

My main interest is 'classical' music typified by R3. For decades I would go to BBC performances, then hear them later via R3. So aimed at a system which when I used them at home got as close as I could arrange to what I'd hear at the same performance, live.

However for rock/pop I tended to want what I liked best. That often meant using tone controls. So what! I've not found that decent tone controls "smudge the sound", if decently made. You can easily enough set them flat or have a bypass option.

Friend used a 33/303/ESL57 setup (had more money then me ) and also used tone controls, because he always wanted a heavier sound from classical and band jazz. Still sounded enjoyable. That's the point. I don't listen toi Hi-Fi to check my ears or 'win a contest'.

Do what you enjoy.

 

[herb]

main interest is 'classical' music typified by R3. For decades I would go to BBC performances, then hear them later via R3. So aimed at a system which when I used them at home got as close as I could arrange to what I'd hear at the same performance, live.

Me too up to recently!

 

[vesuvian]

I expect the time will come when we can easily “measure” our hifi in the home by attaching electrodes to our heads and then we can discuss some really worthwhile measurements.

This is the ONLY measurement that matters - particular regions of the brain are activated when listening to music and every individual will respond differently according to their own musical sensibility. So, even assuming that the lab technicians are measuring the right things (which by and large they're not), their results are completely useless as sound quality indicators.

To return to the OP's original post, clearly most of us CAN hear the differences that cables make and there are obviously mechanisms at work which - shock horror - science has yet to discover. If the measurement nerds were truly scientific in their approach to audio then they would accept this.