There may well be differences in cable sound, but if you have to do multiple run testing and then do a statistical analysis of the results to see if cable A really sounds better than cable B because the difference, if any, is on the fringe of audibility, then I have to wonder, is such a barely detectable improvement worth it? Seem like much ado about nearly nothing.
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Is cable sound real? A more holistic approach trying to track it down.
- Thread starter KSTR
- Start date
OP
KSTR
Major Contributor
- Thread Starter
- #42
Full ack. We must distinguish between the personal perceptual reality which, by definition, is real no matter what caused it, and hearing facts, which is what is left when any biasing is removed to best effort. Once we have established some hearing facts (within the limits of statistical significance), it starts to make sense to look hard for the reasons for it, which should be real differences in the signal as other causes have hopefully been ruled out.
Also note that listening experience and dedication is very important here, IMHO it makes no sense to invite the Average Joe to this kind of tests, enyoing his MP3's on his $200 compact HiFi. As for that, one of the listeners in the mentioned simple blind test in that german forum who runs his own small cable company could identify and blindly name(!) the Vovox cable by one of its "signatures", causing somewhat "blurry bass". Of course this is statistically irrelevant but it still could be an additional data point. I for one have witnessed people with much better hearing than myself in certain regards, pinpointing issues in an instant. One of those people is my former boss, Mr. Klaus Heinz of ADAM(formerly) and HEDD(today), a passionate audiophile and experienced speaker designer. Another was an experienced old-school recording engineer, being half blind, this guy blow me off my socks with regards to what is possible to hear and identify in a dense mix. Finally, we have our dear host Amir who went through a lot of blind test training to identify lossy codec artifacts which many people will probably never be able to pinpoint.
watchnerd
Grand Contributor
So is the hypothesis that:
a) you will identify some phenomena that is reliably audible in DBT/ABX test that isn't explainable by existing LCR models?
Or that:
b) all phenomena are expected to be quantifiable via existing LCR values and you're just establishing a 'threshold of audibility' for skilled listeners?
a) would be an interesting breakthrough
b) is just quantifying something already known (skilled listeners can sometimes hear low level, but real, things normal people can't)
That's essentially the same point that I try to make about the idea of choosing an amplifier based on how you like the sound. Even if one intentionally engineers amplifiers to have a particular sound, why the hell would you? EQing your system via amplifiers is just incredibly inefficient and complicated when you can just...you know...use EQ.
Most of them still design amps and match parts by ear and use measurements as guide. Market for 'colored ' sounding equipment is massive. Tubes, hybrids, turntables..not everybody likes or caters to 'audibly transparent'.
Barely detectable improvement worth it (to you)? Those two words make the difference. Is buying a 10k omega to tell time worth it to you? This applies to every product category from shoes to TVs, furniture, handbags even mobile phone cases, its all the same. Serving the same function with made of similar material with intangible value proposition to the buyer. To prove or disprove fancy pretty cables is like disproving rolex does not keep time better than a tag, and a tag like wise to a lum tec and so on..comes back to personal taste and passion for a hobby. Completely subjective and intangible. And if you want to stick with 10 buck cables with 2/5/10K audio gear for the most logical and scientifically accurate reasons under the sun, all respect to you.
watchnerd
Grand Contributor
I have no problem hypothetically spending 10k on a watch.
Hence my user name.
But spending 10k on an Omega is just lacking in imagination.
Off topic, i dont own an omega too, rolexs triplets breitling and tag. I would like to buy an omega one day though. What watches you like?
watchnerd
Grand Contributor
None of those.
I'm more into over-engineered tool watches and vintage.
At this moment, I'm wearing a Sinn.
I think simple measurements with some low impedance IEMs and impedance varying Focal headphones will show enough differences.
solderdude
Grand Contributor
Will averaging multiple measurements of the same cable help (just like the square-wave plots KSTR did) ?
In that case, how to deal with clock drift.
Will there be a master clock for PB and rec ?
Headphone cables is another nightmare because of the common return wire and Ohmic losses in the connectors and amp.
There are so many cables to measure.
I also wonder if the 'cable with the only correct geometry' was indeed sent to @amirm
We need to differentiate between speaker, interlink, balanced interlink and headphone cables.
Not to mention the effect an RCA cable can have on different cartridges and phono pre-amps.
In that case, how to deal with clock drift.
Will there be a master clock for PB and rec ?
Headphone cables is another nightmare because of the common return wire and Ohmic losses in the connectors and amp.
There are so many cables to measure.
I also wonder if the 'cable with the only correct geometry' was indeed sent to @amirm
We need to differentiate between speaker, interlink, balanced interlink and headphone cables.
Not to mention the effect an RCA cable can have on different cartridges and phono pre-amps.
charleski
Major Contributor
Just remember that the cardinal rule of statistics is “shit happens”. It’s common to use a significance value of p<0.05 for tests like these. A result like that is suggestive, but it also means that 1 time in 20 you will get such a result purely by chance. The key factor here is whether any positive results can be replicated.
It always annoys me when someone comes up with what looks like a positive result on a blind test (there have been a couple for amplifiers) and then touts this as a final answer that ‘proves’ subjectivists are right. No. All it does is indicate that there might be something worth looking at here. Getting a scientific result requires replicating the test, using a different setup to ensure that there isn’t any unforeseen factor confounding the result. Yes, it’s hard, too bad, science is hard. The significance level required for detecting the Higgs Boson was 5-sigma, p<0.0000003.
OP
KSTR
Major Contributor
OP
KSTR
Major Contributor
- Thread Starter
- #55
Any input from members who are well-versed in statistics and scientific approach is welcomed. I'm just a lowly engineer with only little knowledge in those realms.
Same is true for any backup checks of method and execution, that why I came here for this, after all.
All I can say is that I want to try keep things practical, for a start.
----
I hope to have the first set of recordings of XLR cables in analog loopback ready tonight.
OP
KSTR
Major Contributor
- Thread Starter
- #56
Clock drift is not an issue as the DAC and ADC are in one single device fed from the same internal independent master clock (as it's USB asynchronous isochronous mode).
Gain drift is, and there block-averaging will help reduce it (though @pkane might implement correction of slow gain drift in DeltaWave, also he might be able to undo simple -- and microscopic, in this case -- linear transfer function differences coming from different cable capacitances and different time-of-flight values).
Heavy block-averaging has its own problems, though. It also reduces loosely correlated random/stochastic effects, while not as much as uncorrelated noise but still to significant amounts so those might go unnoticed unless we also inspect the noise distribution, sample per sample. An example for this is the excess current noise from some resistors, this can best be isolated by inspecting the noise distribution in the residual (compared to an equal but less noisy resistor), where one can find a modulation of the noise with momentary signal level.
charleski
Major Contributor
The most important thing is to ensure that replication is possible. Document everything, and try to ensure that the setup doesn't rely on factors that might be difficult or impossible to reproduce.
As several people have suggested, if this produces any positive result it will most likely be restricted to a subset of 'skilled' listeners who are sensitive to variations that pass most people by. So you really want to keep track of whoever is doing the tests and be able to invite people to do follow-up comparisons.
[Personally, the idea of devoting time to train yourself to discern differences that will just make you dissatisfied with your hifi seems incredibly perverse. But I'm willing to concede there may be professional situations that call for it.]
yes we need yet another cable tread on the internet as much a hole in the head .
"Is cable sound real? A more holistic approach trying to track it down."
Even the title is slightly biased ? try track what down now ? There is an built in assumption that there is something to track down or explore further that is not explained by simple basics as sane LCR values proper shielding and good mechanical design of connectors .
You can of course explore the endless hole of pathological high end design in both cables and the components themselves.
And find combinations that would surprise , but they are also explained by basic knowledge from 1930 .
charleski
Major Contributor
It would be a good idea to include a ‘sanity check’. Obtain at least one cable that’s objectively, and audibly, inferior: highly inductive, or something like that. You could either fabricate this yourself or stare at one of Audioquest’s more expensive models for a couple of minutes so it breaks.
This will mean that the test contains at least one presentation that should elicit a positive discrimination and provide a baseline.
This will mean that the test contains at least one presentation that should elicit a positive discrimination and provide a baseline.
