|   Cumulative effects | entrope
Jul 10, 2002 7:30 AM | | Suppose you make a system change that causes a physical change to the signal but an inaudible difference in the sonics of system.
Suppose you keep making these type changes?
Shouldn't the results eventually become audible because of an accumulation of physical changes to the output? Or am I missing something technical. I am (obviously) not an engineer nor do I play one on TV or in online discussions.
Finally, if an accumulation of inaudible changes to the output can eventually become audible in the system sound -could cabling be the final tweak in the accumulation of changes that makes the changes audible?
Go ahead and beat me now. |
| re: Cumulative effects | FLZapped
Jul 10, 2002 9:02 AM | | Theoretically, yes. Just like the way Arthur-Anderson kept the books.
-Bruce |
| re: Cumulative effects | A
Jul 10, 2002 9:10 AM | | If you keep adding up many inaudible distortions to the signal, yes, eventually they <i>may</i> add up to something audible, when compared with something that has <i>none</i> of those distortions. The problem with what you are thinking about is this: an inaudible addition to your current setup will not make your current setup sound any different, so wires that inaudibly alter the signal will not change the sound of your system.
To put this another way, if we compared three systems, two of which are so close to the same that there is a noticeable difference, but the least possible, then if we compared either of these to a third system that was half way in between, you could not distinguish between this third system and either of the first two systems, even though you could distinguish between the first two systems. You see, there is something called a 'least noticeable difference', or the amount of difference it takes for something to be noticeable. Anything less than that difference cannot be distinguished from the original. The same idea, by the way, applies to colors. There is a least noticeable difference that people are able to distinguish in shades of a particular color. Anything less will not be visible to humans. So there can be three colors, let us call them A, B, and C, such that A and C are different by the least amount that is visible, and B is halfway in between. A looks the same as B and B looks the same as C, but you can, when looking very closely under good light (assuming you have good color vision), see a difference between A and C. |
| re: Cumulative effects | mtrycrafts
Jul 10, 2002 9:42 AM | | A made a very good point here. UNless you compare the two complete system, you will not hear that under threshold addition. After all, you would be testing the cable against another cable, if I understand you correctly. So, it is the incremental differences you would be judging audible or not.
Tom Nousain published an article
Nousaine, Tom "To Tweak, or Not to Tweak?" Stereo Review, Jun 98, pg 79-81.
While it is not published in JAES, it has merits in spite of its drawbacks. Compared two systems, same speakers,DBT but not rapid switching, so one's memory is poor. But, this is how most would do it at home too, listen to one system, disconnect everything and connect the other one.
Your premis is testable though.Dr Floyd Toole's place could do it :)or CRC in Canada:) Or, just prefer one system over another and I am off your back :) |
| re: Cumulative effects | mtrycrafts
Jul 10, 2002 9:44 AM | | b Or, just prefer one system over another and I am off your back :)
Tell me you like it because of its looks, design, cost, flexibility :) |
| re: Cumulative effects | wanderingbob
Jul 10, 2002 10:40 AM | | I read an excellent analogy once that compared a theoretical audio system to a window made from multiple layers of glass.
If your goal with an audio system is to reproduce the recorded music as accurately as possible, then the original music is like a breathtaking view, say of the Grand Canyon, through an open window. Seeing the view directly is like listening to a live performance: you are receiving the unaltered, original information.
Each component in your audio system (the recording, source, interconnects, preamplifier, amplifier, speaker cables, speakers, listening room) is like a layer of glass between you and the view. The very best audio components are like transparent panes of glass: you experience the music very much like the it was originally performed, with very few added colorations or distortions and very few omissions.
My idea of these kinds of audio components are Krell electronics, Magnepan speakers, and a custom-built listening room, for instance. No, I do not own and can't afford this equipment. (but I'm hoping someday...)
With this analogy, you can see that your view of the Grand Canyon would only be as good as the dirtiest layer of glass allowed; if one layer was dirty and all the others were perfectly clear, your view would still be obscured. Similarly, in an audio system the music reproduced by the system can only be as accurate as the "dirtiest" (least accurate) component in the signal path. Of course, multiple colorations or distorations caused by multiple components can make the reproduced music even less accurate.
Whatever the accuracy of your audio system, if a relatively minor change improves or replaces the "dirtiest" component in the system so that component is "less dirty", then the accuracy of your reproduction of music will be improved.
To give a concrete example, if you had a "mid-fi" audio system where the least accurate component was, for example, a "mid-fi" receiver, then replacing the interconnects or speaker cables - a classical "tweak" - would probably not improve the accuracy of your audio system. If you had a high-quality audio system where the interconnects or speaker cables were the least accurate component, then replacing that component would probably improve the accuracy of your audio system.
In the final analysis, the listener themselves is functionally a part of the audio system. So if you can objectively measure an improvement in the accuracy of reproduction of music in your audio system, but it's not audible to you, then you may very well be the least accurate component in that system! I expect that this would only take place with high quality equipment, quality recordings, and a well-designed listening room.
Bob Gardner
wanderingbob@yahoo.com |
| Interesting - | entrope
Jul 10, 2002 3:23 PM | | I could well be the weakest link in my system. This thought had not occurred to me. HMMMMM.
The analogy was very helpful as was the least noticable difference from "A".
But if the tweaks to two nearly identical systems are removing distortion from one system and the system is more transparent (assuming I am not the component that needs upgrading) it seems that the least noticeable difference could be reduced (systems sound more different) and that cables would play a part.
In my original post I was considering all types of tweaks that have possible minute but inaudible effect (cones, isolation platforms, polarity reversal, raised speaker wire, voodoo dolls, etc) that would in total produce an audible difference. |
| Interesting - | A
Jul 10, 2002 4:28 PM | | You state:
] But if the tweaks to two nearly identical systems are removing distortion from one system and the system is more transparent (assuming I am not the component that needs upgrading) it seems that the least noticeable difference could be reduced (systems sound more different) and that cables would play a part.
The "least noticeable difference" is the least amount of difference that you could possibly notice. This cannot be changed. Two systems can be made to be more different than that, and, typically, systems with different speakers are more different than that. But the "least noticeable difference" is a constant. (For any particular parameter, it may be different than other parameters; e.g., a difference of a certain amount at 1 kHz may be audible, but the same amount of difference at 15 kHz may be inaudible; however, the amount of difference in amplitude at 1 kHz at a particular volume that is audible is a fixed and unchanging amount. Distortion of various types may be similarly analyzed.)
You state:
] In my original post I was considering all types of tweaks that have possible minute but inaudible effect (cones, isolation platforms, polarity reversal, raised speaker wire, voodoo dolls, etc) that would in total produce an audible difference.
First, the thing in question must make an actual difference before it will add anything to your cumulative change (think about the voodoo dolls; I could mention other examples, but I feel like offending as few people as possible at the moment). And then they must add up to enough of a difference to be audible, which may not be the case. For example, one could change the speaker wires to get an inaudible improvement, and also change interconnects for an inaudible improvement. However, it still is very possible that the improvements are both so small that, when added together, they will still be inaudible. And in the case of wires, I don't think that ordinary ones add enough distortion for you to ever get rid of enough distortion by replacing the wires (again, with ordinary wires of normal length; if the wires were 10 miles long, then there would likely be very significant alterations to the signal with an ordinary wire, and then concern about the wire may very well be appropriate). With competently designed equipment, the place where the biggest improvements can be found are in upgrading the speakers, as all of them produce very significant alterations to the signal, generally well above the amounts needed to be audible (this can be shown by measuring the qualities of a speaker to find out its characteristics, and one can find that those distortions are audible by artificially adding that distortion to an audio system, comparing normal playback to playback with the added distortion). (The acoustics of the room are also highly significant to the sound of the system, so changing the characteristics of the room may make very audible differences.) And if the speakers are inefficient or a difficult load, then the amplifier's limitations may come into play. Most electronics are nearly perfect (as far as human hearing is concerned, when operated within its design limitations; e.g., the amp not driven to clipping, etc.), so the upgrades you get from them will get you small improvements (again, when comparing both within the limitations of both, though obviously some amplifiers have far fewer limitations and are therefore better than others in many real world situations). But for you to get an improvement that is audible, you must replace something that alters the signal so much that the added distortion is audible, AND what you replace it with must be so much better that the improvement is audible. This will be the case whether we are talking about replacing one thing or a group of things; you must replace enough stuff such that what you replace has been audibly altering the signal and what you replace it with must be so much better that the difference b |
| continued | A
Jul 10, 2002 4:29 PM | | This will be the case whether we are talking about replacing one thing or a group of things; you must replace enough stuff such that what you replace has been audibly altering the signal and what you replace it with must be so much better that the difference between them is audible.
I should also mention that, obviously, you could change something that makes an audible difference by adding more distortion, and it may sound subjectively good to some people, even though the result is less like the original recording. For example, I owned a tube system many years ago that made Jimi Hendrix sound like Mozart (an exaggeration, but I hope you get the idea); everything had a 'smoothness' to it when played in that system. Well, it was audibly altering the signal (probably by adding distortion and diminishing the treble, but I never measured what it was doing). And that alteration was not displeasing to some who heard the system. I enjoyed listening to it, even though I recognized the fact that it was altering the sound. And if one prefers that, one may call it 'better', but it was less like the original signal. Some 'high end' equipment may be designed to do such things, and some may judge it to be 'better' or more 'musical'. In the case of normal wires of normal length, I suspect that changing the wires to something that does more to the signal will be the only way to make an audible difference, since ordinary wires alter the signal so little. |
| I think this is what he ment... | LordElrond
Jul 10, 2002 8:56 PM | | Hello,
I think what he ment is that if he does alot of little things that are almost negligble they would eventually add up to somthing that is.
Example: 1.Lets say he put equiment on better stand
2.Improved wires
3.Bought a conditioner
4.Bought a fan to blow on amp to keep cool...
5. etc
6. etc..
So the gentleman does the 1st thing..but can't here a differnce, then he does #2 same thing...and so on.
However if he did all these at once he would have heard a differnce...If only minor.
Or for #'s sake lets say that on a hearing scale(imaginary scale for sake of argument) that you needed to improve sound by at least 1 to HEAR a differnce. If you improved sound by 1/6, you wouldn't hear any differnce, or 2/6, but if you did enought little things to get to the magical 1, you would eventually Hear a diffence from when you started.
(To digress...It's like watching a 5yr old grow every day...you'll never notice she's getting taller, but if you measure her say in a month, she might have grown a 1/4 of an inch, but it took all those little days inbetween to make it noticable..)
Anyway that's how I see it...My thinking my be flawed, but It's always fun to discuss...
Mark. |
| least noticable difference | entrope
Jul 11, 2002 4:14 AM | | How can this be a constant if it is related to perception of the person seeing or hearing something (colors/sounds)?
I may percieve a difference between colors that you cannot because of some physical attribute not possessed by you. Or are you saying that the least noticeable difference is a constant related to that person and each persons is different? |
| least noticable difference | LordElrond
Jul 11, 2002 6:08 AM | | I'm saying that if "YOU" the listener could percieve such sounds. Obviously everyone will be diffenent( if just from normal hearing loss and damage)
But the bottom line is if you got the money and time...Do it..it makes you feel better about your system, and gives you somthing to do( and brag about).
That's what HT is for I guess...Messing with it!
Mark. |
| least noticable difference | A
Jul 11, 2002 9:10 AM | | Yes, it can be different for different people. Still, that does not alter the fact that the amount of difference in, say, a 1 kHz signal that you are able to notice is a constant (well, okay, your hearing may deteriorate over time, but it is basically a constant). And there is a best humanly possible, though saying precisely what that might be may be difficult, though it is clear that certain things are beyond human abilities (e.g., the ability to hear 10 MHz is well beyond human capabilities). Certainly, whenever someone claims to hear something that no one has ever demonstrated an ability to hear, it is good to have a bit of healthy skepticism about the matter. And not merely because someone may lie, but because someone can be mistaken about what is influencing them. People are complicated things, from a psychological point of view, and what one seems to perceive may not really be as it seems (the most common example, perhaps, is that of 'seeing a lady sawed in half'; hopefully, one has never actually seen that happen, but has seen something that looks like that has happened). Here are a couple of links, one to a discussion of double blind tests and why they are necessary, and the other to a variety of links, some of which deal with the psychology of perception (e.g., the Placebo Effect):
<a href="/crforum?14@@.ef875dd">A "Explanation of Double Blind Test" 3/30/02 2:46pm</a>
<a href="/crforum?14@@.ef8cc43">A "Wire links." 7/10/02 9:30am</a>
Every now and then, someone objects to discussions of human psychology when discussing what is audible, but one should never forget that all such discussions are about human perceptions, and human psychology is obviously relevant to that. |
| Yes, No, Maybe | sam9
Jul 12, 2002 3:26 PM | | I have read a piece by an amplifier designed of some repute who claims that he designs SS amps that measure THD+N as low as 0.0001% with other measures of sonic quality in the same league. He says right off that that is a couple or orders of magnitude below what is audible (~0.01%), but that since noise and distortion is cumulative he wants to be sure it's not HIS work anyone can point fingers at.
I think there is also some pride/ego in pushing the limit od test equipment, too. |
| Yes, No, Maybe | mtrycrafts
Jul 14, 2002 7:45 PM | | b I think there is also some pride/ego in pushing the limit od test equipment, too.
Yes, that is what engineers like to do :) |
| Yup, Sure Are! | Jon
Jul 14, 2002 7:04 PM | | What everyone else seems to be ovberlooking, whether through ignorance, or sheer deception, is that the audio signal IS NOT STARTING OUT PURE!
It has already been through the sonic mill known as a recording studio.
Audio cable naysayers often like to point out that the signal has already been through dozens of op-amps, miles of wires, and hundreds of coupling capacitors in the recording studio, through the mixing console, the equalizers, etc..
Gee, guess what? These ALL tend to use up the distortion and noise overhead that may exist for "just noticable differences". In other words, the studio recording process is using up at least some of the 'distortion overhead' that does in fact exist, and this process may leave very little error margin left over for the home playback event.
Then, we get to the home playback system, and might find that relatively small changes may now take us above or back below the threshold that was almost exceeded during the recording process.
There is also the issue of audibility, and just how relevant these JND's that get thrown around really are.
First of all, they are population averages, and do not express what a trained or exremely sensitive listener may be capable of detecting.
Second of all, they are for relatively static and simple signals, these being primarily continuous sine waves, with the most complicated test signal being such a sine wave overlayed with a broadband noise signal.
There is actually very little data on just how these simplistic JND's relate to what we might be able to hear with music.
Oh, the cable naysayers like to say that since these are simple signals, that anything we could hear with music would be higher still, not less, but the fact is, they have VERY LITTLE such data to go on, and a lot of it was done some years ago.
When pushed to supply even ONE such reference, mtry and most all of the other naysayers have YET to provide any such reference.
The fact that despite all of our "understanding" of the human ear/brain mechanism, we can not make audibly perfect codec algorithms should tell you that we still don't know all the magic ingredients that go into what we CAN hear.
Then there is another factor, one that Richard Heyser exposed, that of an experience threshold, with hysterysis, that one has to cross, before we notice something, and then, the conditions have to go back some ways before we can no longer notice the newly discovered effects.
See:
"A View Through Different Windows", Audio magazine, Feb. 1979
AND
"Catastrophe Theory and It's Effect On Audio", Parts 1 thru 3, Audio magazine, March, April, May 1979
The point being, that there is more to it than just JND's and thresholds of hearing, much more.
Jon Risch |
| Yup, Sure Are! | mtrycrafts
Jul 14, 2002 7:53 PM | | Jon, your citations are so old that it is pre-history. You are so fond of saying that old citations are not good, yours are even older, so they must be worthless then, right.
b These ALL tend to use up the distortion and noise overhead that may exist for "just noticable differences".
Bull. There you go again, speculation, 'may.' Where is ytour evidence? Do you even know what JND is? Doesn't appear to from this speculated example. LOL.
b and this process may leave very little error margin left over for the home playback event.
There you go speculating: 'may'
Do you ever have any data, for a change?
And the rest of your tome is full of 'may,' and empty of any evidence. Why is that, Jon? |
| Yup, Sure Are! | Jon
Jul 15, 2002 9:53 AM | | When I refer to "out of date" citations, it has been primarily concerned with 20 + year old cable listening tests, where the only things being tested, were essentially two differnt sizes of zip cords, on equipment with substantially less resolving power than what is avaialble today. I would say that these fctors are relevant to rendering these "old" listening tests completely worthless.
On the other hand, the articles by Richard Heyser are more along the lines of thought pieces, where he explores a different way fo looking at things, and explores the whole process of listening and of detection.
These were revolutionary in their time, and are still relevant today, as a good idea never gets dated.
If you understood enough technical information, perhaps we could talk about how noise and distortion can add up, how signal abberations are going to increase as it passes through more and more less-than-perfect devices. This is no mystery or any kind of unknown, competent audio engineers have known about these kinds of things for decades.
You might try reading the Heyser article, then perhaps you could make an attempt to discuss it intelligently.
Jon Risch |
| Yup, Sure Are! | mtrycrafts
Jul 15, 2002 2:55 PM | | b You might try reading the Heyser article, then perhaps you could make an attempt to discuss it intelligently.
LOL.
I do have all 4 articles in hand. I think it is you who should try to read it, better yet, offer evidence how it relates to audio. Obviously the author is lost in the science fiction aspect of his beloved reference, "Flatland, A Romance of Many Dimensions," 1884, that he tries to use as some sort of cross application in audio. LOL.
He should try to offere some evidence with real world DBT listeing, LOL.
And, what has happened since his article in 1979? Nothing. No further evidence from him, or you, LOL.
b This is no mystery or any kind of unknown, competent audio engineers have known about these kinds of things for decades.
You are too funny, Jon. Yet, you have zero evidence for audibility. LOL. Keep on speculating as did the author, both of them, E.A. Abbot included.
b These were revolutionary in their time,
LOL. No, Jon, great imaginations, no different from what you try to offer 23 years later. |
| Yup, Sure Are! | Jon
Jul 15, 2002 6:34 PM | | mtry,
All I can say is that when you start to try and dismiss Richard Heyser, a true visionary and a bonified genius, you are going to lose the respect of a lot more people.
Even I am amazed at your intolerance, lack of understanding, and utter foolishness regarding Mr. Heyser's work.
I couldn't possibly begin to explain it to you mtry, not any more than I could adequately explain a rainbow to a blind man.
Jon Risch |
| Yup, Sure Are! | mtrycrafts
Jul 15, 2002 8:20 PM | | b All I can say is that when you start to try and dismiss Richard Heyser, a true visionary and a bonified genius,
Oh, he may be a genius but he certainly missed the boat about the correlation of his articles to audio, especially when he just expressed his speculations how it may correlate into audio.
b Even I am amazed at your intolerance, lack of understanding, and utter foolishness regarding Mr. Heyser's work.
What work would that be? The one in Audio? Really? I didn't see any data in there, just trying to explain multi dimentions having some sort of correlation in audio. He missed the boat.
Perhaps you have a better citation from him that appeared in some Journals? Hopefully relating to audio.
b I couldn't possibly begin to explain it to you mtry, not any more than I could adequately explain a rainbow to a blind man.
No, Jon, you will never be able to as there is nothing to explain but speculations. You nor he has offered any data that relates to audio. |
| He is rolling in his grave | mtrycrafts
Jul 15, 2002 8:33 PM | | you trying to use his Audio paper to support your jibberish.
Nothing in it. |
| Yup, Sure Are! | FLZapped
Jul 15, 2002 4:08 PM | | Wow, wee, what a rant. Sure makes all your wailing about 1.3dB at 20kHz seem awful insignificant now, doesn't it.....
-Bruce |
| I am amazed | mtrycrafts
Jul 15, 2002 8:50 PM | | b Oh, the cable naysayers like to say that since these are simple signals, that anything we could hear with music would be higher still, not less, but the fact is, they have VERY LITTLE such data to go on, and a lot of it was done some years ago.
When pushed to supply even ONE such reference, mtry and most all of the other naysayers have YET to provide any such reference.
That you have yet to offer any data as to the use of music to establish JNDs or, better yet the threshold of hearing. Why is there no such data, Jon? Could it be because your speculations about the usefulness of music is not worth discussing among professionals? But you wouldn't know that.
But, all the known JND studies and threshold data uses single, simple tones as they are the most sensitive to detect. |
| Irrelevant to home audio. | Pat D
Jul 17, 2002 6:53 PM | | There may and probably are lots of changes to an audio signal in a recording studio, most not due to the wire. Purchasers cannot compare the recording with the actual performance or even the live mike feed. What we have is a CD, DVD, SACD, LP, tape, etc.
What happened in making the recording has no relevance as to the utility of special interconnects or speaker cables for home audio. We have recordings of various formats. The usual question here is whether special interconnects and cables make an audible difference in our audio systems. Bringing in red herring arguments does nothing to establish your contention that they do. Very poor logic, Jon. |
| Irrelevant to home audio. | mtrycrafts
Jul 17, 2002 9:06 PM | | b Very poor logic, Jon.
Has he ever use good logic? That will be a first :) |
| Irrelevant to home audio. | Pat D
Jul 22, 2002 4:37 AM | | Well, I just looked at his formal argument, irrespective of the empirical claims, and found it made no sense. |
| |
