[Kees]

Quote:

Originally Posted by rodbac If everything observed fits into a well-established paradigm, which these reports do, there is no reason to advance some alternate explanation, ESPECIALLY when that alternate explanation breaks down so quickly. If you don't think you're hearing placebo, and that 20KHz signal is actually being altered over one adequately sized piece of copper versus another, then show everyone a scenario where placebo doesn't work and you've got an unexplained phenomenon for which your "we don't know everything about this"-scenario would apply. While a reasonable scientist would be the first to insist that science never knows everything, there ARE some things it knows very well, and the behavior of an electrical signal is one of them.

I am pretty shure we don't know everything there is to know about electrical signals.
In extreme situations electrical signals behave very unpredictable. Ohm's law is not reality: it is a working model. And it is not complete.
Like Newton gave us a working model for gravity and mass that worked pretty good for a long time (and is still used by many). Only it was not reality.
Einstein proved that.
I don't say Ohm's law is wrong. I just say Ohm's law is probably not all there is to it. It is just the best we've got right now.

 

[rodbac]

Quote:

I am pretty shure we don't know everything there is to know about electrical signals.

You might be right- so now all you need to do is show an EE a situation in this realm they can't predict accurately and you're in business.

I'm being dead serious when I tell you you could very well get a Nobel Prize for it.

 

[romistrub]

First off, let me establish my credentials. I'm a second-year nanotechnology engineer at the University of Waterloo in Canada. I know a little bit about sound, a little bit about electricity, and a lot about quantum mechanics. Now that I've got my lameness out of the way... let me have a go at this.

Ohm's law is one of the luckiest laws that exists to this day. It is a very basic law based completely upon Maxwell's equations for the movement of charge through an electric field. The "real" Ohm's law... which has nothing to do with Ohm... is stated as "Current Density = constant * Electric Field" where total current flowing through a conductor is equal to the "Current Density" as a surface integral of a cross-section of the conductor.

In layman's terms... a single electron will accelerate in an electric field proportionally to the strengh of that electric field.

What this article is trying to do is idealize the system into a simple RCL system, which is a great theoretical framework for doing real experimental tests. Unfortunately, that is first year university mathematics, not even close to the caliber of research being conducted at Sennheiser for example.

The author makes assumptions such as the non-existence of errant electric and magnetic fields, which is completely ridiculous considering that our houses are filled with countless wireless and wired devices creating electromagnetic radiation and interference patterns all throughout. It's like trying to design a car for perfect roads, perfect weather, and a perfect driver. See how well that car performs in real-world conditions.

I agree that the burn-in time for solid-state electronics is something beyond my scientific knowledge (I'd rather say that than say there is no such thing). I cannot imagine a conductor or even a system of conductor/insulator materials could change with otherwise controlled usage.

*EDIT* I Just thought of something: While there may not be changes within a conductor due to running a current through it over a long period of time, there may be physical changes to the geometry between conductors. There are a large number of tiny gaps (or rather, huge gaps when deailng with a small enough scale) in the signal path of most solid state electronic devices occurring at wire-meets-circuitboard transitions and similar locations. While an electric signal is passed through the "gap" the metal atoms of the conductors react to it and vibrate slightly. This will occur until the surfaces have a maximized connection geometry. Think of it like placing the last pieces on a 95% complete puzzle (crude analogy, I know). It is quite feasible to simply vibrate the table until the last 10 pieces fall into their respective places "automatically"... even though this may take years (again, this is a very crude analogy).

I definetly agree that the analogue-is-better business is a thing of the past. 192KHz/32bit recordings do enough justice to the ears to keep me happy. (Note to everyone: there is no such thing as true "analogue" recordings. Every recording is limited in sample rate and bit rate by the physics and structure of the media on which it is recorded)

I would go into more detail but I'm getting bored of writing... so I'm sure you're getting even more bored of reading. Anyways, don't consider me to be arguing my points "authoritatively"... I have about 1/4 the education this guy has... just in a completely different field. Consider me a skeptic

.

 

[romistrub]

Any feedback on my crazy ideas? hehehe

 

[rodbac]

romistrub, you can make all the speculations about errant fields (et al) you want, just like everyone else, then you'll get the same challenge:

Find a single scenario in the realm we're discussing where these laws you're trying to diminish don't predict the behavior of the signal perfectly, and you might get a Nobel Prize to go with your bachelor's degree.

Oh, and what does this even mean:

Quote:

Unfortunately, that is first year university mathematics, not even close to the caliber of research being conducted at Sennheiser for example.

This subject has been mundane for generations. Trust me- Sennheiser isn't wasting time doing *any* research on getting their anticipated 20KHz (max) signal from your amp to their headphones.

How best to reproduce that signal once it gets there, though, almost assuredly.

PS: It's good, encouraged in fact, to be skeptical as you imply you are. Nothing wrong with that at all.

[edit]

There's really nothing more I can say on this that hasn't been said already. If anyone wants to believe there are unknown forces at work on your 20KHz signal, please go ahead and do so.

I'm out.

 

[barkas]

@kees: that's probably enough

. But I don't believe that cat5 cable copper is in any way superior to power line copper or any other copper in any cables. It is used to transfer digital signals with a pretty good crc and resend if anything is lost: They are not optimal, but just good enough for the job while being reasonably cheap.
The better cables (cat5e/cat6/cat7) don't have any other copper either, as far as I know, but the shielding is substantially better, which won't help you. So using cat5 for speaker wire is probably a good idea for me, having a lot of it that I don't need right now, and having a source for it that costs me almost nothing, but probably not a good idea for everyone else, because that stuff is not substantially better than speaker wire of sufficient diameter, costs more, and you have the hassle of combining a lot of strands into one contact, which will not improve the contact any.

@romistrub: digital burn in sounds pretty far fetched. There are things like electromigration, but that is not an interesting effect in audio chips since the traces are much too big here. I believe this effect and vibration effects are much too small to be hearable or even measurable. It is probably a better idea to be interested about your amp and what chips and topology is used than in those effects.
Stray electromagnetic fields could change the signal of course, but they would have to be very strong and near, and you could always use shielded wires.

 

[Kees]

Quote:

Originally Posted by barkas @kees: that's probably enough . But I don't believe that cat5 cable copper is in any way superior to power line copper or any other copper in any cables. It is used to transfer digital signals with a pretty good crc and resend if anything is lost: They are not optimal, but just good enough for the job while being reasonably cheap. The better cables (cat5e/cat6/cat7) don't have any other copper either, as far as I know, but the shielding is substantially better, which won't help you. So using cat5 for speaker wire is probably a good idea for me, having a lot of it that I don't need right now, and having a source for it that costs me almost nothing, but probably not a good idea for everyone else, because that stuff is not substantially better than speaker wire of sufficient diameter, costs more, and you have the hassle of combining a lot of strands into one contact, which will not improve the contact any.

It is not the quality of the copper that does the trick. It is the combination of many different strands that are twisted and braided (though all of them seperately insulated off course).

It minimizes the risk of signal interference / distortion.
It offers ample capacity.

The skin effect of the signal traveling through separate SOLID THIN wires that are constantly crossing each others path actually causes the signal to utilise the capacity far more efficiently.

The skin -effect is the effect that the current mainly travels through the outside "skin" of the wire. In this design there is just a lot more "skin" the signal can use. The skin-effect pushes the current towards the outside of the wire-construction, the braiding & twisting keeps guiding it inside again, thus utilising all of the capacity to its max.

This is the short version of the theory. I can explain in more detail if you are interested.

24 leads are just about the max one can handle with ordinary spades or bananas....

Based on this I just tried a couple of things and found a very satisfying result.
That's all there is to it realy....

 

[rodbac]

Kees, "skin effect" is completely irrelevant at audio frequencies.

Here's a good writeup- and please check the math if you don't want to trust him:

http://www.audioholics.com/techtips/...ect_Cables.htm

 

[romistrub]

Quote:

Originally Posted by rodbac Find a single scenario in the realm we're discussing where these laws you're trying to diminish don't predict the behavior of the signal perfectly

Okay, fair enough. I suppose that minute gaps in the wire could be considered as capacitors, etc.. I guess the most important thing to note is that we're talking about making a signal audible. There's no point addressing issues that only affect the electronics on a micrometer/nanometer scale, because I doubt that it would affect the electron flow enough to mess with the signal enough to be audible anymore than a fly hitting your car causes a noticeable dent.

Quote:

Originally Posted by barkas I believe this effect and vibration effects are much too small to be hearable or even measurable.

You may not be able to hear the difference, but it is most definetly measurable. If we can use electron microscopes to see shapes of carbon rings, we can definetly detect a fluctuation of even 10eV in the signal... if we wanted to.

As far as my crackpot theory being far-fetched... it is for audible signals... but it definetly created measurable changes in the signal.

Anyways, thanks for setting me straight!

 

[Kees]

Quote:

Originally Posted by rodbac Kees, "skin effect" is completely irrelevant at audio frequencies. Here's a good writeup- and please check the math if you don't want to trust him: http://www.audioholics.com/techtips/...ect_Cables.htm

It does not say it is irrelevant to audio signals. It says it makes a VERY small (3%) difference in RESISTANCE value. Which is totally irrelevant to my arguments to use the Litz wire design the way I did.

Edit:
To make that a bit more clear:
I am talking about the "width" of the signal path. Due to the skin -effect the width of this path is directly proportional to the surface of the wire. The difference of this signal path between "one thick solid wire" and the same wire split into "24 equal thin solid wires" (separated!!!! Cross-over between the wires would spoil the effect)) is then directly proportional to the (cumulative) circumference of the wire(s). This yields a signal path that is 4.8989795 times larger ("wider") for the "24 thin wires"-construction.....

 

[Kees]

@barkas
We can discuss wether a difference in the signal path is measurable in the signal or not.
We can discuss if a certain person can hear a certain difference or not.

But I can, and will, not discuss what you BELIEVE to be audible or not

As far as I am concerned anybody can believe anything he or she likes.....
Because everybody agreed that it cannot be proven that anything is inaudible, your belief is in this discussion irrelevant.

 

[Kees]

@rodbac
What I see as an example of the many differences between theory and practice of "Ohm's law":
http://sound.westhost.com/amp_design.htm
Time and time again, in practice it turns out that measurements are dfferent than could be expected from theory.

It is an interesting read in its own right too.

 

[barkas]

Quote:

Originally Posted by Kees @barkas We can discuss wether a difference in the signal path is measurable in the signal or not. We can discuss if a certain person can hear a certain difference or not. But I can, and will, not discuss what you BELIEVE to be audible or not As far as I am concerned anybody can believe anything he or she likes..... Because everybody agreed that it cannot be proven that anything is inaudible, your belief is in this discussion irrelevant.

There's not much to go on beyond belief, I fear. What is audible to you may not be audible to me. What will probably be audible to most is as far as science can go here.

 

[Patrick82]

Is that website a joke (superhero costume)? They compare this hobby with a restaurant...

I don't think any experienced audiophile believes what they say.

 

[Vul Kuolun]

Quote:

Originally Posted by Patrick82 Is that website a joke (superhero costume)? They compare this hobby with a restaurant... I don't think any experienced audiophile believes what they say.

If "experienced audiophile" is somebody who removes a correction circuit from his state-of-the-art headphones altering the sound ceveral decibels at 3000Hertz on a range of 3 octaves, and blames the change on a stupid cable, you might be right.

Maybe you wanna tell us what the joke about this really is.