Things that make you go HMMMMMMM!!!????

[esldude]

http://18.7.29.232/bitstream/handle/1721.1/46225/41567257.pdf?sequence=1

 

This link is in another thread. Is an MIT EE student doing tests of interconnect to see if any measurable artifacts could explain why they might sound different. There are a few nits to pick with this or that detail, but pretty thoroughly done.

 

To summarize, measuring S/N, THD, IMD, phase, frequency etc. etc. all the levels appear so low even under extreme impedance for interconnect use that none of it could explain how they could sound different. His comparisons are with a couple inexpensive RS wires and Monster from their inexpensive to very expensive interconnects. Only channel to channel crosstalk might be close to being audible according to the measurements. Everything else is flat and clean at such very low levels over extremely wide frequency bands (by audio standards) they just don't appear they could make any remotely audible difference.

 

So if interconnects sound different why? What is going on?

 

I say this as an owner of some moderately pricy cables that I do think make a difference. But this certainly gives one pause as to whether or not it is real.

 

I am sure many will say "who cares, if I hear it then it is so, does not matter why". But I am more curious than to let it go at that.

 

[wdomeika]

Since then the methods, tools and effectiveness of marketing hype have improved exponentially resulting in seriously expensive interconnects and cables that now unequivocally sound better than old plain old copper wire.

 

Cheers,

 

Bill

 

[bzr]

...interference , the different types of connectors, what the connectors are made of, it's really not hard & doesn't need to be long winded.

That being said, there are some things around here of late that make me HMMMMM!

 

[esldude]

If you look at the results in the link, it doesn't appear to be material or connectors. I don't suppose he made any direct tests of interference suppression though he looked at construction and shielding. I know materials and connector construction is how they are made differently. It just doesn't appear those provide any meaningful differences in measured ability to pass an audio signal.

 

They still may be why things sound different, but still leaves us wondering why?

 

[mayhem13]

For analog line level interconnects, if the connection is sound meaning proper and clean, then there's only one thing tht might affect say and amp or preamps abilities and that would be impedance. Slight changes would impact the performance but whether or not it's actually audible is another story entirely. As for RF and EMI, following simple rules of not crossing over or running parallel to voltage wiring is a must.

 

The recent discussions over digital transmission cables is much more alarming. There's no middle ground here. They're either dropping data which is CLEARLY audible ot they're not. Suggesting otherwise that there's an audible effect in the analog signal chain is simply absurd.

 

 

[Julf]

I definitely see where you are coming from, but I try to keep an open mind, while still being a strong believer in scientific skepticism. I know conventional engineering theory mostly supports the "either it works or not" view, but there seems to be a significant group of people claiming to hear a difference, so in all fairness, I do think that it is important to be open to the possibility that there are things going on that our current engineering models don't account for - maybe interference or noise in the cables cause recoverable errors that still affect timing in less-well-designed USB interfaces?

 

Anyone remember TIM (Transient Intermodulation Distortion) back in the 70's? Until professor Otala pointed out the potential timing problems with large amounts of end-to-end feedback, everybody was dismissing people who claimed transistor amps sounded bad by pointing out the extremely low distortion figures and linear frequency response compared to tube amps, and claimed there was no factual support for the claims of bad sound.

 

At the same time, this is also an area where human psychology can interfere rather badly with objectivity, so I think there is a strong burden of proof required (sort of in the league of ESP and Bigfoot) to state that established scientific theories are wrong. So that is why a lot of us insist on things like double blind testing.

 

Of course, it is up to each one of us to decide for ourselves what to believe in, and how to use our money. If you think (digital) cable A sounds better than cable B, and that makes you happy, then please go ahead - but don't state that what you think is the truth, without providing something more to substantiate it than "I can hear the difference" or "see the article in ".

 

Julf

 

 

[4est]

IMO, it is foolish to assume that everything can be distilled down to a simple macro impedance without losing some micro information along the way.

 

I mean that literally and figuratively.

 

Different cabling does sound different on some equipment. That is a fact, not a fallacy. Only when we can accept that, and honestly look for reasons why will we begin to understand the issues at play.

 

[esldude]

Well, I am asking for reasonable suggestions for "issues at play". If we assume something is a fact, when the only measurable rigorous facts at hand suggest otherwise, we might well be looking for a cause that in fact does not exist. You won't ever get anywhere like that. So where is it we should honestly look?

 

If extremely low noise from interconnect, no phase changes that can matter, no distortion and no restricted frequency response are occurring then what is? BTW some of the tests in the link were into extreme impedance situations beyond almost anything you find between consumer audio equipment. So it doesn't seem to be an impedance issue.

 

[CG]

It may not be an impedance matching issue, but the impedances can have something to do with it. Since the load end is usually very high impedance for an interconnect, the interconnect could be thought of as a capacitor in parallel with the load. That's one simple outlook.

 

So, the details of that capacitor would matter, would they not?

 

http://www.national.com/rap/Application/0,1570,28,00.html

 

http://waltjung.org/PDFs/Picking_Capacitors_1.pdf

 

http://waltjung.org/PDFs/Picking_Capacitors_2.pdf

 

 

 

 

[Jud]

To summarize, measuring S/N, THD, IMD, phase, frequency etc. etc. all the levels appear so low even under extreme impedance for interconnect use that none of it could explain how they could sound different.

 

The very same could have been said of CD players when first introduced. No one could come up with relevant measurements that accounted for anything less than virtually perfect sound quality. Later on, measurements of things like jitter began to account for sound quality less than perfection, and for differences between players.

 

Julf has also alluded to Dr. Otala and the use of feedback in solid state circuits.

 

These are just a couple of the numerous examples of two trivially correct facts:

 

- At any given time, even very good scientists and engineers don't know and consequently haven't reduced to measurement everything that goes into the sound quality of audio equipment and cabling.

 

- Some of what has been reduced to measurement is not widely known to the public either because it is "hidden in plain sight" in academic sources using academic language not easily understandable by laypeople, or because it is closely held as proprietary information by equipment and cable designers.

 

My take on this is that until folks who flatly say "X cannot possibly make a difference," or, for that matter, "X is clearly what makes the difference" provide citations to the literature and/or academic qualifications, I'm not even going to begin to credit such arguments. And when I say "citations to the literature," I mean not only measurements, but clear and cogent arguments that those particular measurements are the only ones relevant to sound quality.

 

Fortunately, our most knowledgeable contributors - those who actually have the academic qualifications and audio engineering experience - seldom tend to speak in such absolutes. (In fact, Gordon Rankin has mentioned his frustration at *not* being able to reduce sound quality completely to measurables among an identified set of variables, and the difficulty, perhaps impossibility, of clearly understanding the interactions between all relevant variables.) Of course this doesn't mean good audio designers lack any idea of what makes good sound quality. It just means they are smart and knowledgeable enough to understand they don't know everything.

 

[4est]

We do not listen to sine waves, and most testing is done with them. I do not intend to argue about this, but the idea that everyone only "thinks" there are differences is ludicrous at best. For instance, the rise times in a composite wave can be nearly that of a square wave. That indicates to me that there is a need to be capable of at least an order of magnitude beyond what we hear to not have an effect within the audible range.

 

My point was simply that I do not think we will get anywhere if we have to prove we can discern differences. There are many things that we do not know how to measure or the act of measuring swamps the system to the point where measurement in that fashion is meaningless.

 

Are there issues that you would honestly consider reasonable? That is a rhetorical question btw.

 

[esldude]

Something like TIMD of Otala or jitter yes, those kinds of issues I would consider reasonable if there are such. But what are they?

 

The issue with sine and squarewave and music is a red herring. All of that is well understood the difference isn't that one uses test signals that aren't music.

 

Saying we can't get anywhere if we have to prove we hear differences is most certainly not getting us anywhere. By proving what you can and cannot hear you can narrow down what might be causing the genuine audible differences. There are definite limits to what can be heard. If you insist on infinite hearing abilities then you can never get anywhere.

 

[4est]

If it were so well understood, we wouldn't be having this discussion and you wouldn't feel the need to address a nitwit such as I. As you said, "Things that make you go HMMM.".

 

I've never said my hearing or anyone's is infinite. I also do not feel that our measuring abilities are infinite either. I obviously do not have anything more to add to this...

 

[esldude]

The Walt Jung articles show why not all capacitors are equal. Quite measurable differences. Basically not all capacitors, especially electrolytics perform as the ideal capacitor should over all frequencies. The materials in most cables come close enough it shouldn't be an issue and it doesn't measure like one.

 

In the link of the opening post, Mr. Cooper, the MIT student, does two different simulations based upon measured values of the cable to see what is expected. One uses Matlab and I forget the other. He then runs the actual tests. Capacitance is part of those simulations. His measured results differ from simulated results almost not at all. The difference is truly trivial. So I don't think capacitance alone explains what is happening if interconnects sound different.

 

[esldude]

Sorry, I am being perhaps too argumentative.

 

We do have some things in common. I notice you have Soundlabs M3 speakers. I have Aura's which use the same panel with different cosmetics. Surely we have some other things in common.

 

Never called you a nitwit nor imply such.

 

[4est]

No worries, I am not offended and I called myself the nitwit, not you.

 

I am just making the suggestion that there is plenty of anecdotal evidence to suggest that cable CAN make difference that we do hear. The jury is out on whether or not it is a positive (meant more accurate or what have you) difference.

 

With that said, one attracts more flies with honey. Your answer most likely will not come from the cable skeptic, but from the cable believer. By which I mean that the non believer will not provide much evidence, and therefore not lead anywhere. The believers, on the other hand, by omission or commission may lead you somewhere...

 

[esldude]

Yes differences aren't always for the better. Or may sound better though a coloration. Certainly I will not argue that more transparent is always more enjoyable.

 

I think tube amps are a very good example. They seem to be better. Seem subjectively more transparent, seem like a clearer view to the source and almost always are more enjoyable. But it is a fortunate group of colorations that make that so. It is so, it isn't figment of imagination, it is real, it is measurable it even will pass blind testing.

 

I don't mind that, I would just prefer it be understood as such. Then we can have less expensive though good and powerful transistor amps that can be pretty transparent. We could intentionally and selectively color the sound for enhanced enjoyment at the line level or digitally more effectively than having to use the actual full size tube power amp.

 

And almost everyone knows you never want truly flat 20-20khz response from your speakers in your room. It simply isn't enjoyable with virtually all recorded material.

 

[Mark Powell]

Are there any? I cannot recall anyone, on this thread or anywhere else, say that cables make no difference. Maybe my memory is defective. In my experience they certainly do, but find that expensive ones are not in general superior to low-cost ones. And that I will defend to the last bullet!

 

[Julf]

Just to pick some nit - you say that "there are definite limits to what can be heard". Right. But if it can't be heard, does it matter?

 

[esldude]

Just to pick some nit - you say that "there are definite limits to what can be heard". Right. But if it can't be heard, does it matter?

 

Not to how something sounds. If we determine limits of what is audible and push all the reproduction to lower levels then it becomes a non-issue. Which seems to conform to the idea of high fidelity when sound is listened to by humans.

 

Can matter in non-sonic ways. If you can get a highly efficient switching amp to equal the sound of class A monster monoblocks then you can get equivalent sound for less energy. Less operating costs, less heat in the listening room in summer etc. etc.

 

 

 

[Jud]

Of course the trouble is that it is often not easy to be certain whether something affects the sound. I recall that Spectral, a manufacturer with a sterling reputation (and wonderful equipment, as I can attest having been a very happy owner for 18 years) gave what seemed to me as a layperson a fairly cogent explanation for why their amps had reasonably flat response out to near 1 mHz!

 

[CG]

(Original message deleted. Life is far too short for this!)

 

[esldude]

I believe Spectral, among other reasons wanted very gentle response roll offs for the whole system. They wanted first order slopes. 2nd order slopes they thought would result in resonances that among other things could reflect into the audio band. One need only listen to their equipment to find they know something as it is top notch good sounding gear.

 

Is 1 mhz or later 2 mhz bandwidth a necessity for equally good sound? I don't know, maybe well enough done 200 khz is enough (you do usually need that in an amp to make sure you don't audibly alter 20 khz). They took an extreme approach to be sure it worked. No doubt backed up by listening and using their own recordings. They wanted the cables to control a nice even roll off of system response by having the active portion capable of more than enough bandwidth. Have used one of their pre-amps and think it probably the best or among them I have ever used. Have owned a power amp and can confirm it had response all the way to 1 mhz and beyond as claimed.

 

Their gear was/is unusual in having all the supposed virtues of solid state with none of the once common drawbacks. It seems entirely transparent as well. So whether their approach is overkill or not those guys at least know one way to get it done right.

 

 

 

[1audio]

Having designed the first and second generation Spectral products I'm not sure what you are talking about. I'll check with Keith and see if anything has changed. The response was flat to at least 500 KHz on the amps and 1 MHz on the preamps. I always avoided peaking the response to extended it for good reasons (even though I think Shunt Peaking is very valid in some applications). The low frequency response was always extended to DC except for the phono section. I believe the overall concepts have not changed.

 

I have always contended that extended response sounds better if for no other reason that the execution quality it forces on you.

 

Also (perhaps this is where the roll off question comes from) cascading a number of roll offs in different products produces a low pass or high pass filter with uncertain qualities, possibly with a peaked response. The classic fix is to stagger the roll-offs if they are close (not really good) or move them so that a single one dominates.

 

[esldude]

Glad to hear from you Demian. I knew you were with Spectral, but didn't realize you were the designer.

 

I think it was an interview with Rick Fryer where he was describing the idea of managing roll offs and I think he was referring to the result of the whole system. Obviously I cannot argue with the fellow that designed it as to why.

 

Those were some terrific designs. I owned a DMA 50 connected to Quad ESL63 speakers. I think the phono stage of the pre-amp is the best I have heard, the whole pre-amp actually. The preamp I got to use is one you fellows made up for Jim Smith that could handle microphone inputs. Have also known people with later products too. Top notch stuff.