Analysis of Paper on Measurements of RCA Cables by Kunchur (Video)

[anmpr1]

When the paper came out, I wrote three objections in the comment section of the paper. All three were rejected by the moderator and discarded...

Pathetic. But why am I not surprised? You'd think it would be different. After all, it's the AES, not Superbest Audio Friends. Are you a member of AES? That would be doubly disappointing behavior on their part, if you are.

On the other hand, it probably doesn't matter. With ASR, your work reaches many times more influential people than a blog post on an obscure thread could ever reach. My guess anyway is that most AES members are likely reading ASR--probably more than they read the Journal.

 

[amirm]

Yes I am a member of AES. Only members are allowed to post comments.

 

[pseudoid]

I am a scientific skeptic but consider all 'peer-reviewed' papers worthy of at least a read before the file/or/flush decision.
So, I must admit that I watched the video and ended up going down some serious but tangential rabbit holes that were educational.

This Kunchur paper skool'd me that EMI stands for 'Electro-Magnetic Induction' [].
Yet, I finally learned that an RCA cable with a directional arrow is a shielded cable but the shield is terminated at the 'source' end ONLY.
Who am I to character-assassinate the writer and/or our local ASR' peer-review committee!

I would like to thank @amirm, Kunchur and the committee!

 

[Kyle / MrHeeHo]

The craze on YouTube over this study reminds me of just how sensationalist reporting on scientific papers often is

 

[Spkrdctr]

Indeed. "We accept any and all feedback provided it is glowingly positive!"

That is true with so much of our society today. If you are not 100% positive you are considered to be negative and a troublemaker. Critical thinking is so rare in todays society. I am hopeful on ASR threads as I see some fellow critical thinkers are still in the world. It always gives me hope. Plus Amir's reviews and videos show critical thinking and I like that too.

 

[krabapple]

True, I should have started my sentence with "But". But more fundamentally, I thought that was an odd response, like "you must be new here, audibility is proved by blind tests." First, simply because ASR never performed any blind tests, and as far as I know haven't done much promotion for scientific findings done by blind testing.

Of course it has. For example: the work of Toole and Olive on loudspeakers is cited repeatedly.

The rest of your post is really just a series of similarly dubious claims masquerading as facts.

 

[PeteL]

Of course it has. For example: the work of Toole and Olive on loudspeakers is cited repeatedly.

The rest of your post is really just a series of similarly dubious claims masquerading as facts.

Thanks, can you post me a link on Olive blind test studies that enlight us on audibility metrics? What is it that I claim which is dubious?

Edit, I think you are talking about speakers preference rating.Sure yes but speakers do sound different, you are trying to assess what is the preferred one. Cables don’t sound different, that’s a totally other challenge in term of statistics. Blind test won’t tell s much unless you test a very large population. Null test are way more useful.

 

[Cars-N-Cans]

Hope all you want. Despite repeatedly being shown that his "papers" and "research" are incompetent, he continues, makes excuse after excuse, publishes even more in the same vein, and continues in a way that will benefit him personally. It's transparently dishonest and, frankly, more than a bit disgusting.

Getting back to my earlier comment I can only assume then that this translates to his field of expertise as well, which is unfortunate. I’m sure if we had the level of knowledge in astrophysics as we do here in audio the silhouette of the iceberg below the proverbial waterline would be quite visible. He probably just does a better job of it when he knows important people are looking.

 

[krabapple]

Sure, you are right, but my point is that blind testing have a very limited scope in science, if any.

You can't possibly mean *science* generally. The use of blind protocols is too widespread in science for that to make any sense.

Blind testing of video codecs… sound ominous

Only if you don't understand what 'blind testing' actually means.

 

[krabapple]

Thanks, can you post me a link on Olive blind test studies that enlight us on audibility metrics? What is it that I claim which is dubious?

I'm not going to buy into what *you* think matters (ill defined 'audibility metrics' in this case) when you seem so oblivious to the role of blind testing in the study of hearing , not to mention use of blinded protocols generally in science.

Edit, I think you are talking about speakers preference rating.Sure yes but speakers do sound different, you are trying to assess what is the preferred one. Cables don’t sound different, that’s a totally other challenge in term of statistics. Blind test won’t tell s much unless you test a very large population. Null test are way more useful.

Yes, null testing, done right, is pretty definitive evidence that the two devices under test are audibly indistinguishable. So are other types of measurement. That;s no excuse for making ridiculous dismissive claims about blind testing. Go educate yourself. Get Zwicker and Fastl from a library and read the section on sensory testing.

 

[SIY]

Getting back to my earlier comment I can only assume then that this translates to his field of expertise as well, which is unfortunate.

That's your assumption. I have no idea why you'd make that leap, I certainly wouldn't, but you're certainly entitled to your own opinion..

 

[voodooless]

Only if you don't understand what 'blind testing' actually means.

Or actually have a sense of humor…

 

[lc6]

"Garth compared AudioQuest's new Thunder cable ($700) with AC cables from other companies priced up to $18,000, culminating with the AudioQuest Dragon ($4000).
It was fascinating to hear how different cables accentuated or attenuated various aspects of the sound, such as the attack on the acoustic guitar's treble, the sibilance on Waters' voice, or the space around voice and guitar. Not surprisingly, the AQ cables presented the best-balanced sound, which Powell ascribed in part to better rejection of the RF energy that surrounds us in the connected world."

"One time, at a different CES, using a well-recorded trumpet disc, Garth Powell demonstrated the "directionality" of a green ground wire. I'm serious. It was ridiculously easy to hear. [...] All AudioQuest "Mythical Creatures" Interconnects use what the company calls "Solid Perfect-Surface Silver (PSS)" conductors; they use the same material for RF-draining. The dielectric is FEP (fluoropolymer) air tubes, and there are other features you can find on the company's website along with a "white paper" explaining the tech behind each one."

From the cited "white paper":
"So, would there be any benefit in eliminating (as best we could), the cable’s characteristic impedance? Absolutely! Could it be done? Yes — the issue is markedly reduced by eliminating the cable’s dielectric constant via 100% electrostatic shielding. ZERO-Tech (no characteristic impedance) is a technology that I developed for AudioQuest’s Storm Series of AC power cables [...]"

The formula for characteristic impedance is:

View attachment 227042

So, if "no characteristic impedance" means that it is zero, there are two possibilities:
R + jωL = 0
or
G + jωC = ∞

The first would require zero resistance and zero inductance. The second would require infinite conductance or infinite capacitance or both. Neither is physically realizable. And of course, as shown in @amirm 's video, characteristic impedance of a cable has no effect over the full range of audible frequencies, let alone at the 50 or 60 Hz mains frequency.

What is also puzzling: "eliminating the cable's dielectric constant" implies it is zero. But by definition, the lowest value of it is 1 (in vacuum), and every material's measure is relative to that. Decreasing the dielectric constant also decreases capacitance, which means the C term in the formula above decreases, which makes Z-naught bigger all other things being equal. Am I right?

To continue on the wild theories of electricity according to Garth of AQ, just a couple of excerpts from yet another paper found in the Theory & Education section of the company's website:
"If you use the AC service tap from your utility, many power amplifier manufacturers will recommend a dedicated 20-amp outlet (North America 120V) or 16 amps (220-250V). [...]
The problem with this in many systems, or with the power distribution products as I’ve listed, is that, as the number of discrete dedicated AC lines used for a system increases, the further one veers from a single-point AC ground reference and a bad hum from ground current loops may occur. [sic; grammar]"

The only way this could be the case is if those dedicated AC lines were daisy-chained. But they all have individual runs (i.e. hub-and-spoke), including the ground wire, to the service panel. Per the national electric code (NEC), ground connections may only be shared among different circuits in a raceway (a common conduit). And the ground connection, as explained in @amirm 's educational video, is only for safety, i.e. when the user-touchable part of the load becomes energized.

"One important technique for those with either split-load (North American homes) or a professional building (three-phase service; most of the world) is to make certain that all audio/ video AC service taps are on the same phase! You will need to work with your electrician to have that wired to your AC outlets or confirm that is already the case. The voltage differential (difference) Line to Neutral and Neutral to AC Ground will vary substantially (phase to phase) at the service panel, so the likelihood of hum from a ground
loop increases dramatically in that scenario."

First, the power supply to US homes is single phase. So there are no multiple "phases." Second, there is only one Neutral and Ground connection (to the center tap of the power transformer) shared between the two hot Line connections. So if there are two Neutral connections that have any voltages (let alone different voltages) at the service panel to the Ground, that would indicate a wiring fault. Now, a voltage between Neutral and Ground at the wall receptacle (as opposed to the service panel) is a different story. And using the "same phase" for won't help because voltages across Neutral connections to different receptacles are independent (since loads and neutral wires are independent).

 

[SIY]

Blind test won’t tell s much unless you test a very large population.

That depends on what question the experiment is designed to answer.

I'm not sure why you classify experiments by what their basic controls are.

 

[Phorize]

I agree and stand corrected on the EM spectrum. However that does not take you out of the quandary that "audible differences" are testable. In the end, all human actions are dependent on physics (e.g. me typing now). But if human audible differences were to be testable then ASR is only dealing with 50% of the problem (electronics etc. but not people).

Are you saying that there is a reason to doubt the position that over 100 years of research on electrical transmission of sound has brought us to, being that linearity is what matters, best judged with reference to noise and distortion? What is that reason?

 

[PeteL]

I'm not going to buy into what *you* think matters (ill defined 'audibility metrics' in this case) when you seem so oblivious to the role of blind testing in the study of hearing , not to mention use of blinded protocols generally in science.



Yes, null testing, done right, is pretty definitive evidence that the two devices under test are audibly indistinguishable. So are other types of measurement. That;s no excuse for making ridiculous dismissive claims about blind testing. Go educate yourself. Get Zwicker and Fastl from a library and read the section on sensory testing.

OK, tell me. Instead of calling me uneducated, enlighten me, that's why I'm here for..Let's stop generalities and speak about the subject at hand. I am not saying to not use blinded protocols in science, I am talking about audibility of cable. If ten people can't distinguish the difference between 2 cables. Does that demonstrate that there is no audible difference between the two cables? What about on a sample of 100 people? What about if you have 1000 and 1 individual spot differences and spot them repeatedly and rigorously. What does that mean? You can't see, even without making calculations, just intuitively, that proving by blind testing a "hypothesis", scientifically demonstrating by other mean in this case, that there is a statistical problem is this reasoning? That it is not a proper scientific method to prove that kind of thing and that it's a different statistical problem than demonstrating there is a trend when assessing which type of frequency response listeners tends to prefer in which case it's a totally valid method?

 

[PeteL]

That depends on what question the experiment is designed to answer.

I'm not sure why you classify experiments by what their basic controls are.

That was my point yes.

 

[SIY]

OK, tell me. Instead of calling me uneducated, enlighten me, that's why I'm here for..Let's stop generalities and speak about the subject at hand. I am not saying to not use blinded protocols in science, I am talking about audibility of cable. If ten people can't distinguish the difference between 2 cables. Does that demonstrate that there is no audible difference between the two cables? What about on a sample of 100 people? What about if you have 1000 and 1 individual spot differences and spot them repeatedly and rigorously. What does that mean? You can't see, even without making calculations, just intuitively, that proving by blind testing a "hypothesis", scientifically demonstrating by other mean in this case, that there is a statistical problem is this reasoning? That it is not a proper scientific to prove that kind of thing and that it's a different statistical problem than demonstrating there is a trend when assessing which type of frequency response listeners tends to prefer in which case it's a totally valid method?

"There's a pony in there somewhere!"

40 years of flamboyant claims and no evidence. It's pretty safe to conclude it's all a bullshit fraud.

 

[SIY]

That was my point yes.

No, you said "Blind test won’t tell s much unless you test a very large population." This is not a generally true statement.

 

[PeteL]

No, you said "Blind test won’t tell s much unless you test a very large population." This is not a generally true statement.

Correct but you did not follow the whole discussion. It was to answer "Audibility is proved by blind testing" and I made the precision, that I am not saying blind testing generally has no scientific value. You can't remove a statement from it's context. But it is "generally" a true statement that those statistical methods are only valid if the population tested is also valid. It depends on what you are testing.