Request for help: brains needed!

[Richard Black]

I make this request in all humility and in the spirit of co-operation. We all know that there is a shortage of published tests demonstrating cable differences. By that I mean differences that are plausibly responsible for sonic changes from cable to cable. The holy grail, I suggest, would be any kind of non-linear distortion, i.e. frequencies present in the output that weren't in the input.

I'm trying to do a few tests on this. My question is this: what tests do people recommend to demonstrate real differences from cable to cable, over and above simple (and small) amplitude and frequency response changes? Before you reply, I'd like to add three things:

1) Let's stick to interconnects for the time being. They're easier to work with!

2) If you're a cable 'believer', what tests (IN THE AUDIO BAND unless you've got a VERY strong reason why not) will show up differences?

3) If you're a cable skeptic, what tests will once and for all convince the waverers (better accept that true believers will never be converted!) that cable 'sound' is illusory?

I have personally tested basic THD and IMD-type stuff to better than 120dB tolerance. I have looked unsuccessfully for 'jitter'. I have nulled cables against one another (with correction for amplitude and phase due to simple LCR filtering effects) to better than 85dB, using real music programme. I have used synchronous averaging to check for signal-consistent distortions to even lower levels (such averaging doesn't really help look for noise modulation or interference-related effects). I have even tried putting a microphone in front of the speakers to make sure I am testing the _whole_ system, though that's damnably difficult to get any sense out of. Other people on this forum and elsewhere have tested for the same and other parameters with similar lack of success - as far as I know.

What's missing?

Richard

 

[Az B]

Our local audio club went to a local independent testing lab for cable testing field day a while back. The main tests were for differences in resistance, inductance, and conductance, as well as full audio frequency tests. The cables tested ranged from reptilian and Nordic gods to the cheapo interconnects that you get when you buy a $200 CD player.

The tests clearly showed little, if any difference in the audio spectrum. In fact, the cables tested nearly identically, with the high priced cables having some degree of positive improvement in frequencies beyond human hearing.

It was a very enlightening trip. But there's still two camps in the club. Those that accept the results, and those that are still wondering why they can hear a difference.

I think it all boils down to whether you are a person of logic and reason, or a person of emotion and intuition. There's little that can be done anymore to prove either side right or wrong.

Personally, I've always attempted a balance between the right and left halves of my brain, but even then it's hard to stare into the face of cold reason and come up with anything different than what the facts state.

 

[jeffsg4mac]

Richard, I suggest you start by going though all this One of Gene's favorite pastimes is debunking cable myths. There is a ton to digest here.

 

[gene]

Richard;

Cables DO NOT cause non-linear distortions, but don't take my word for it, check out what Bruno from Phillips has to say about it along with his test data from one of the worlds most accurate devices for measuring distortion in the following article: Debunking the Myth Of Cable Distortion

Assuming you are testing two well designed interconnects such as a Quad Shield Coax or Shielded twister pair cable from belden, they should sound indistinguishable provided they are well terminated, well shielded, and capacitance isn't excessively high. Only poorly designed cables are sonically distinguishable.

If you want to do a cool experiment, how about treating the first reflection points in your room with absorption material, and treat the wall behind the listener with diffusors, then blindfold a bunch of listners and have them listen with and without the added room treatment. I bet you will get much more correlation in this test, then you would playing with a few interconnects

 

[Richard Black]

<<Cables DO NOT cause non-linear distortions>>

This absolutely matches my scientific upbringing and my own, don't-trust-it-until-you've-seen-it-with-your-own-eyes experimenting. However, I have hypothesised that non-linear distortions could quite plausibly occur, and probably do in some instances. Here's one: if a cable picks up a goodly bit of line-frequency hum from somewhere (you can imagine a scenario - unshielded parallel wide-spaced pair, crummy transformer nearby) the result in a real loudspeaker will be intermodulation between music signal and the hum. It's a two-point failure, sure, but results will vary between cables. That particular one I have tested for and found to be lower in level than I can measure however hard I try in realistic instances with real audio cables, but you get the drift, I hope.

Richard

 

[Richard Black]

PS - I have indeed read - most assiduously - Gene's and Bruno's articles refered to. And looking again at Bruno's reminds me that I have also test microphony myself, including such sensitive tests as fairly whacking a cable, in the presence and in the absence of an applied signal. With realistic source and load impedances, nil result (i.e. comfortably below -100dB).

Putting that number in there also makes me reflect that of course we are in a sense also arguing about minimal audible difference. In the case of any kind of non-linear effect I'll be persuaded to take interest at levels around -100dB. But linear effects? Who reckons they can spot a -60dB (o.1%) level change? You don't normally see it quote like that, of course: a change of about 0.01dB. The air between you and the speakers can change its attenuation in the upper audio band more than that as the room warms or cools by very few degrees, or gets a little more or less humid.

R.

 

[Richard Black]

PPS - I left this forum and decided to spend a few minutes reading some cable bits and bobs on the web. Ended up on the MIT website, where I found one of the best bits of smoke and mirrors handwaving I've ever come across in audio. Strictly for those with a bit of a grounding in electric circuit theory, give yourself a grin at:

http://www.mitcables.com/technology/power7.asp

go to the bottom of the page. The leap in logic is just wonderful.

R

 

[gene]

Here's one: if a cable picks up a goodly bit of line-frequency hum from somewhere (you can imagine a scenario - unshielded parallel wide-spaced pair, crummy transformer nearby) the result in a real loudspeaker will be intermodulation between music signal and the hum.

That is not cable distortion. What you are describing is noise pickup which migrates through the system and then gets amplified in the electronics (assuming the electronics lack adequate filtering). The cable is NOT the mechanism for the distortion and yes in this scenario shielded cables should be used which is what we always recommend.

 

[mtrycrafts]

It was a very enlightening trip. But there's still two camps in the club. Those that accept the results, and those that are still wondering why they can hear a difference.

There's little that can be done anymore to prove either side right or wrong.

.

This is why DBT listening is used to eliminate or zero out human bias
This is the truth test

But, not much will convince a true believer

 

[Richard Black]

<<Quote:
Here's one: if a cable picks up a goodly bit of line-frequency hum from somewhere (you can imagine a scenario - unshielded parallel wide-spaced pair, crummy transformer nearby) the result in a real loudspeaker will be intermodulation between music signal and the hum.



That is not cable distortion. What you are describing is noise pickup which migrates through the system and then gets amplified in the electronics (assuming the electronics lack adequate filtering). >>

Depends how you define distortion, of course (the most rigorous definition is any difference at all, of any nature, between input and output, excepting only amplitude scaling), but it's certainly system distortion in which the cable plays a significant role. There are other examples of apparently trivial 'distortion' in one audio component being transformed into something more serious elsewhere in the chain - 'Alias Intermodulation Distortion' in CD players plus loudspeakers being a case in point. I wrote an AES paper on that, a few years ago.

Richard