[AussieMick]

Are you really sure CEOs are spending their days listening and designing systems? Not really sure what an expert is if you think we have to find golden ears. (sorry, just my wording about how the test pool has to be exceptional). Heck, if we're saying the end all be all test for an audiophile ethernet connection with test pool, why not increase it? The premise is that it's obvious that an audiophile ethernet device eliminates noise and improves audio: but we need golden ears and audiophile components to test? Sure, if these companies are willing to participate in ABX, that's great. But also not sure why it has to just be exclusive to them if the claim is that it's demonstrably clear there's a difference. If it's demonstrable, why not a larger random sample?

Chief designers, not CEO. If you really want an answer then yes, the test pool should be exception. In a world where some don’t hear differences between DACs, we’d better be very careful. All we need is specialised experts to pick it in the test, then we know it’s real and get on with finding a measurement. Or we find out it’s not real and move on.

 

[AussieMick]

Also, a random sample gives the experiment less control. We’re not trying to find out “can a random selection of people perceive a difference”. We’re trying to find out if there is sufficient evidence to warrant a rethink on measurement procedures. It seems we’d have more control over the experiment if our subjects were “expert” level.

 

[Davesrose]

Chief designers, not CEO. If you really want an answer then yes, the test pool should be exception. In a world where some don’t hear differences between DACs, we’d better be very careful. All we need is specialised experts to pick it in the test, then we know it’s real and get on with finding a measurement. Or we find out it’s not real and move on.

Lets get back to practicality: audio companies tend to be hesitant about ABX tests because their profits are about marketing. I'm dubious that a chief designer is some kind of golden ear, compared to other listener. But then again, results with ABX will show you that! You say some hear a difference in DAC and some don't: how much of that is anecdotal stories?

So I'm pointing this out because I'm thinking you're not going to get these specialized experts you'd like. It it then falls to consumers, why should it matter if it's someone who's an audiophile vs someone else? The premise was that this is an obvious difference. The more people you can include...whatever best audiophile or their significant other is higher sampling rate.

 

[Davesrose]

Also, a random sample gives the experiment less control. We’re not trying to find out “can a random selection of people perceive a difference”. We’re trying to find out if there is sufficient evidence to warrant a rethink on measurement procedures. It seems we’d have more control over the experiment if our subjects were “expert” level.

Why? The premise is that this devices obviously improves ethernet connections and improves audio. That seems exactly "can a random selection of people perceive a difference".

 

[digititus]

Lets get back to practicality: audio companies tend to be hesitant about ABX tests because their profits are about marketing. I'm dubious that a chief designer is some kind of golden ear, compared to other listener. But then again, results with ABX will show you that! You say some hear a difference in DAC and some don't: how much of that is anecdotal stories?

So I'm pointing this out because I'm thinking you're not going to get these specialized experts you'd like. It it then falls to consumers, why should it matter if it's someone who's an audiophile vs someone else? The premise was that this is an obvious difference. The more people you can include...whatever best audiophile or their significant other is higher sampling rate.

A positive correlation signal (can hear a difference) will be independent of the sample population (provided they are able to hear), if the testing setup is to be considered valid and reproduceable (blind ABX).

 

[Davesrose]

A positive correlation signal (can hear a difference) will be independent of the sample population (provided they are able to hear), if the testing setup is to be considered valid and reproduceable (blind ABX).

Can you cite examples?

 

[AussieMick]

Yes, but we’re working to a bigger picture than this one product, aren’t we? And the moment the experiment is compromised in any way the “losing” side has an out. Better to make it beyond that.
DACs and anecdotal evidence - well there’s heaps of it on this forum. I’d suggest if a person can’t hear a difference between a Chord Mojo2 and Hugo2, then they’re not likely to hear something more subtle (like the LAN thing) are they? If they’re not likely to hear a difference they really have no business being in the trial because it will skew the data far too much.
I know people who can’t tell the difference between Maldon Salt, Himalayan Salt and plain iodised table salt. Does it mean there’s no difference? Heck no. Should those people be involved in taste tests of subtle differences in salt? Also no.
Hearing is a sense that, like salt, can be trained. If we want to know the real answer to products of this type we need trained listeners.

 

[AussieMick]

Can you cite examples?

if the burden of proof is on the maker of the statement, I suggest the person who openly and repeatedly called this device and the company fraudulent should need to prove it.

 

[Davesrose]

Yes, but we’re working to a bigger picture than this one product, aren’t we? And the moment the experiment is compromised in any way the “losing” side has an out. Better to make it beyond that.
DACs and anecdotal evidence - well there’s heaps of it on this forum. I’d suggest if a person can’t hear a difference between a Chord Mojo2 and Hugo2, then they’re not likely to hear something more subtle (like the LAN thing) are they? If they’re not likely to hear a difference they really have no business being in the trial because it will skew the data far too much.
I know people who can’t tell the difference between Maldon Salt, Himalayan Salt and plain iodised table salt. Does it mean there’s no difference? Heck no. Should those people be involved in taste tests of subtle differences in salt? Also no.
Hearing is a sense that, like salt, can be trained. If we want to know the real answer to products of this type we need trained listeners.

At this point, I don't think we are. Or I should say ethernet products: as audiophile ethernet popped up in the last few years, and it's been ongoing with computer folks saying "ethernet is ethernet" and audiophiles saying "I hear a difference with this". We've been talking about the best way to have a controlled test to actually settle things with skeptics vs audiophile. Now productively, the bigger picture would be a controlled test that's scientific and could possibly lay to rest if audiophile ethernet is valid.

You bring up things like Chord Mojo2 and Hugo2, or that you compared computers vs dedicated streamer....these are all variables different than just a controlled ethernet connection. Seems you're saying if someone hasn't heard a Chord, then they don't know audio and they can't be discerning. I mentioned previously about what your PC or streamer could be is due to processing: I admit that I also find some amazing things with how a device could process and its different output stages. Also lets face it that most audiophiles are males and they have a significant other that will lay down the law. I have had instances where my large speakers were questioned, but then when music or movie track fires, it's "OMG, this is better than a movie theater". For some audiophile reviews on YouTube, they have their significant other give impressions and it's either the same or valid difference.

I really think that if this is so obvious, why not a large sampling pool of "average" folks, "audiophiles", and as you say "experts" (but I'm still skeptical that companies are going to be willing).

 

[AussieMick]

Seems you're saying if someone hasn't heard a Chord, then they don't know audio and they can't be discerning.

Be careful, I haven’t said anything like this at all. Furthermore, it’s something I absolutely don’t believe. Please don’t put words on my mouth.
I don’t think the difference would be “obvious”. I think it’s most likely very subtle if anything at all. I haven’t heard it.

 

[digititus]

Can you cite examples?

Purely from a theoretical standpoint of how to design such an experiment. IF there is a correlation then this should be observable regardless of the sample population. The device is responsible for the change and not the experience or "insert label" of the listener.

 

[AussieMick]

Purely from a theoretical standpoint of how to design such an experiment. IF there is a correlation then this should be observable regardless of the sample population. The device is responsible for the change and not the experience or "insert label" of the listener.

But if the difference is small enough that only an experienced listener can reasonably be expected to spot it then your point doesn’t hold. We’re not talking about something like a drug and a placebo remember.

 

[Davesrose]

Be careful, I haven’t said anything like this at all. Furthermore, it’s something I absolutely don’t believe. Please don’t put words on my mouth.
I don’t think the difference would be “obvious”. I think it’s most likely very subtle if anything at all. I haven’t heard it.

OK, maybe I had a problem with if a person can hear a difference with Chord Mojo 2. Whatever their experience level....the largest sampling rate in this test for ethernet signal is optimal. You previously mentioned only highly experienced folks should participate. If you can get those folks from companies to participate, great. But including as many people as you can is just a better sampling pool.

 

[Mr BubbaHyde]

Bought it 2 days ago , i think zen stream and Lan isilencer are perfect match.
No need to Argue
This is only a US$100 toy, which is a lot cheaper than acoustic revive, jcat, sotm.
I think you should challenge the above manufacturers

Thanks for checking it out, this was exactly the use case I am going to try it in!

Cheers!!

 

[ktm]

OK so noise can cause issues with ethernet. The symptoms are
Packet loss. But normally this isn't the case. It would require
Something noisy next to the cable. A ceiling run of cable laying
On a fluorescent light is a common source. Other than that, data gets
from a to b just fine. Rf would have to pretty strong to cause data loss.
So that leaves noise of lower levels going from a to b or getting picked
up on the cable itself. You should be able to measure ground plane
noise. Check it with the cable plugged in and disconnected.
You don't need golden ears and blind tests for this. We could beat this
to death like that. Until someone shows the lab work it's not going to
Prove anything. You'd expect a blacker background not better highs,
bass, etc.