Ethernet System for Audio: putting it all together

[acg]

2 hours ago, davewantsmoore said:

2 hours ago, dbastin said:

The network is a source of noise (like the AC grid, and RFI).

I think that's actually very unlikely the be true.

 

However... noise is relative.   Nothing is completely "noise free".

 

Also the susceptibility of various audio devices to any "noise" may vary.

 

So do I, but having recently drunk this Kool Aid I can understand why people would think that the network is a source of noise, because of how the network changes made here actually change the sound...like there is less noise in what you hear.  Honestly, I can't make sense of the "why" and talking to people about it is just confusing because of differences in technical backgrounds, how we each describe sound, and the reasoning each person correlates to apparent changes in sound.

[Assisi]

6 hours ago, Nada said:

 

Even these subtle observations are testable with a DBT to see if they are repeatable.  It is possible to clarify if they are associated with physical changes or better explained by the complexities of attribution.

 

Just need two of the same switch.  Put them in the test system for a week, both  running. Do a DBT to ensure they are no different. Then turn one off. Turn it back on. Retest. Easy.

I started my network journey a few years ago with a very inexpensive switch.  I eventually tried an audio switch out of curiosity.  Low level but a small benefit.  The rest is history.  Overtime I have built what I consider to be a serious network setup that for me delivers benefits.  This thread is about building a network. 

 

 

I am not positive to your suggestion.  I comment on my experiences assuming that some posters are interested in what I say.  I am not fussed if the responses from others doubt my experiences.  Your response or that of anybody else to test the validity of my experiences often comes back quickly, do a BDT.  To me the request is just an excuse to cast doubt on my comments.  It does not interest me.  To me it is about the mystery of what is happening and why.  I ask my self constantly the questions of what and why?  I often do not have the answers that may satisfy you or others.  That doesn’t overly matter to me.  I do know though that reasonably frequently something has happened and that is why I comment.   It is all about an improvement in the listening pleasure.

 

 

A few people have listened to my system.  The responses from the few are definitely that I have achieved an impressive outcome.  I do not need confirmation of a DBT.  To undertake a rigorous controlled DBT probably requires serious planning, special setup and acoustic conditions.  I suggest that it is more complex than you outline.  Can you provide an outline of any really successful and conclusive DBT that you have participated in or are aware of?  Have you changed your system because of a rigorous DBT that influenced you?

 

John

[NPC]

7 hours ago, Assisi said:

I responded with an invitation for him to listen to the evidence.  For me the evidence is about the contribution of the overall sum of the parts of my system including the network and not just one component.   It was about listening to my entire system.  It is not about one component, the Melco switch as he suggested.  I construed his suggestion as  superficial. 

 

What is your motive and Why focus on just one component?

 

In the context of this thread and similar ones covering Ethernet, it is entirely apposite to focus on any component(s) that anecdotally made such profound differences in your environment. Why is that? (rhetorical). This is why the focus. I hope this provides you with a frame of reference.

 

Vis a vis the offer to listen to your system "as is", it would be a futile exercise I am afraid, as Alan pointed out. To "listen to the evidence" when we cannot then vary/substitute/change components means one would have no yardstick with which to judge the existing components you have hyped if you are unwilling to make the changes in order to open up and experiment and to gain the consensus of others. Without a little A/B swapping and testing, if all you want is for people to come and listen and say "very nice", then not much has been gained aside from a little bit of appendage waving which has no benefit to anybody whatsoever; let alone  in the context of these labours.

 

I think this may now become circular and at the risk of appearing cynical it may appear to some that this denotes a lack of confidence in whether the components in question would be able to deliver in light of the build-up they have been given if we cannot compare.

 

We'll leave it there until the circumstances change.

 

 

[dbastin]

 

2 hours ago, acg said:

So do I, but having recently drunk this Kool Aid I can understand why people would think that the network is a source of noise, because of how the network changes made here actually change the sound...like there is less noise in what you hear. 

Hmm, great point ... like there is less noise.  It baffles me what else could be the cause.

 

I have assumed it is noise because of things like:

• changing from the SMPS to a low noise LPS on a switch gives an improvement
• Using a Gigafoilv4 which says it reduces noise gives an improvement (no special clocks in that BTW)
• grounding ethernet devices improves the sound, even changing the ground cables does

All these are aimed at reducing noise, and sound like they do so based on the changes other noise reduction things I have done.  It seems like a logical conclusion, but may be incorrect.

 

I wonder if the Kool Aid cable you tried and its propriety bespoke shielding includes things to absorb noise?  That is becoming more common these days - nano things and magic crystals.

[acg]

1 minute ago, dbastin said:

I wonder if the Kool Aid cable you tried and its propriety bespoke shielding includes things to absorb noise?  That is becoming more common these days - nano things and magic crystals.

 

I certainly hope there are no crystals or nano things...or quantum things.

[dbastin]

4 hours ago, Grizaudio said:

 

Happy to try, just not happy to pay without listening first.

 

I think this deep dive is getting a bit much for me.

I might need to wave a white flag and tap out. 

 

I also disagree with what Dave said about "I think any improvement in the signal matters.....  even if it is just academic.".

IMO, Improvements are all relative to what one can achieve elsewhere in the system with the same investment. 

 

If a listener needs to concentrate intently, compare A to B 20-30 times to hear a benefit - I would argue the benefit doesn't exist or it simply doesn't matter. 

For instance, if I could spend $3000 on network improvements v's $3000 on an new amplifier or subwoofer addition, I know which I would prefer. 

If differences are more easily heard, well then a value judgement needs to be made. 

 

ATM, my current streamer uses Wi-Fi and I'm happy with that.

I not against your suggestions, but my next step is to upgrade my streamer and subs.

 

We all have different base lines and priorities.  BTW my 1st suggestion may be only $150, but your current wifi might be better as it is also total isolation.

 

I agree, you need to consider where the best bang buck will come from for you and your system.  At some point investing in the network will be the best value thing to improve your sound quality.  After all, if you stream data, that is the source.

[davewantsmoore]

2 hours ago, Assisi said:

To me the request is just an excuse to cast doubt on my comments.

 

It's actually, completely, the opposite.

 

If you can pick between your "network", and straight through piece of $1/m cable .... clearly and reliably under the conditions I described ....  I will stand up and take notice.  (If I was there) I will expect to be able to hear the same thing for myself (or be somewhat embarrassed).

 

It's the way I would indicate to someone "ok, I'm actually interested, let's see what substance is here".

 

 

2 hours ago, Assisi said:

I do not need confirmation of a DBT.  To undertake a rigorous controlled DBT probably requires serious planning, special setup

 

Not at all.   It requires running one cable.... and then a casual listening session.    With someone to switch the plugs over, and keep record of which is which at what time.   Nothing more, nothing particularly special.

 

There doesn't need to be anything "rigorous" about it.... or any "double" blinding.

It's quite simple really.

 

The reason is doesn't need to be so fancy (is like FN was saying) is that you are not looking to differentiate between "unknown" and/or "tiny" differences.    There needs to be no rigorous process that is fit to "prove" anything.....   it only "disproves" something.   The test will be able to hear the difference (which it should if the difference is clearly different) .... or not be able to tell the difference.

 

2 hours ago, Assisi said:

Can you provide an outline of any really successful and conclusive DBT that you have participated in or are aware of?

"D" BT..... ie.  double blinded is a step further than what is needed.

 

Just typical blinded tests ...... zillions.    I routinely do it with speaker designs, where I have A and B change to a speaker design - say a different crossover (and I may have an opinion on it already, maybe not) ... and I see if I can identify the difference between them (through either casual listening, or ABX, or....)

 

2 hours ago, Assisi said:

Have you changed your system because of a rigorous DBT that influenced you?

 

Define "rigorous"  <shrug>

 

I've chosen many things cos I could clearly hear them blinded..... and I've put less importance on other things, where it was demonstrated that they weren't very audible (or audible at all).

 

 

In these types of tests.... it is not ideal to know I am testing something..... and it is not ideal to know what I am testing.... but this is difficult to avoid.    It is essential to not know which is which out of same A and B ..... and that's about it.   For "relatively large" changes, this is completely sufficient.

 

[davewantsmoore]

9 hours ago, Grizaudio said:

I also disagree with what Dave said about "I think any improvement in the signal matters.....  even if it is just academic.".

IMO, Improvements are all relative to what one can achieve elsewhere in the system with the same investment. 

Of course.

 

If you have a "budget" .... then there are some huge important things to focus on.   Flat frequency response is the king of everything.

[acg]

8 minutes ago, davewantsmoore said:

Of course.

 

If you have a "budget" .... then there are some huge important things to focus on.   Flat frequency response is the king of everything.

 

Network tweaks are high on the audio tree of things to get right...certainly go for the low hanging fruit first.

[frednork]

21 minutes ago, davewantsmoore said:

Not at all.   It requires running one cable.... and then a casual listening session.    With someone to switch the plugs over, and keep record of which is which at what time.   Nothing more, nothing particularly special.

 

There doesn't need to be anything "rigorous" about it.... or any "double" blinding.

It's quite simple really.

 

The reason is doesn't need to be so fancy (is like FN was saying) is that you are not looking to differentiate between "unknown" and/or "tiny" differences.    There needs to be no rigorous process that is fit to "prove" anything.....   it only "disproves" something.   The test will be able to hear the difference (which it should if the difference is clearly different) .... or not be able to tell the difference.

It could be that simple if the result was that it is discriminated successfully.

 

If not ,it is a useless result that shows nothing unless you go back and show that a known similar small difference was distinguishable by the tester/s on that experimental setup.  And as I said, even then not proven but just more likely that it may not be distinguished by the participant/s on a reasonably resolving experimental setup (as you were able to show that other similar level things could be distinguished). But even  that is much more valid than the vast majority of non-academic tests I have seen done. 

 

The fact that it is Johns system and he is used to it is irrelevant as he is no longer "just" listening" to it and the conditions have changed due to the test.  This applies to anyone using their own system that they may know very well. 

 

The level of blinding does not need to be double as you have said.  The protocol will determine what  level is required. If the person changing the cables is out of sight of the participant than double blinding is not necessary.  Of course you would also need to ensure there are not other clues such as the different sound the cables make when inserted in the switch etc.

 

I would really encourage people to try to see if they can reliably distinguish something that they know is definitely measurably different and if they can then go from there with smaller differences. It may not be as easy as many think. Especially when you get a couple of extra people involved to ensure adequate blinding. 

 

 

[davewantsmoore]

2 hours ago, frednork said:

on that "experimental setup"

I do really think you are in danger of over complicating things.

 

If we are talking about difference which is (known to be) big and significant .....   then It's just turning a stereo on and listening to it casually.... and then after a while, doing a thing, and listening again.... and then after a while doing a thing again, and listening some more.    Not doing anything special.... not flicking back and forth .... just sitting around listening to tunes (and writing down some notes)

 

All that is needed is that nobody can see which cable is plugged in / being switched....    it's not that hard.

 

The outcomes are either that:

 

We can hear the big thing... or

The thing we thought was big, might not be so obvious after all.

 

[frednork]

3 minutes ago, davewantsmoore said:

I do really think you are in danger of over complicating things.

 

If we are talking about difference which is (known to be) big and significant .....   then It's just turning a stereo on and listening to it casually.... and then after a while, doing a thing, and listening again.... and then after a while doing a thing again, and listening some more.    Not doing anything special.... not flicking back and forth .... just sitting around listening to tunes (and writing down some notes)

 

All that is needed is that nobody can see which cable is plugged in / being switched....    it's not that hard.

 

The outcomes are either that:

 

We can hear the big thing... or

The thing we thought was big, might not be so obvious after all.

 

Anything to do with ethernet is not a "big thing"

speakers are a "big thing"

[davewantsmoore]

1 hour ago, frednork said:

Anything to do with ethernet is not a "big thing"

Some say it is quite noticable  <shrug>

I'm only suggesting this sort of test for that situation.

[frednork]

7 hours ago, davewantsmoore said:

Some say it is quite noticable  <shrug>

I'm only suggesting this sort of test for that situation.

considering you are well aware that there is no measurement thus far to confirm this "big thing" change I can only assume you are being disingenuous here

[davewantsmoore]

14 minutes ago, frednork said:

considering you are well aware that there is no measurement thus far to confirm this "big thing" change I can only assume you are being disingenuous here

No.   I am not.

 

We have a claim that "there is something significant" (my paraphrasing).

 

We don't know what "audio measurements" of it look like....   we could try to get them .... but then there are also (sometimes) claims that it is audible no matter whether or not it is measurable.    I dispute that (the something audible could me "un-measurable") .... but there is no need to get tied up on that.

 

Any significantly audible thing .... will pass a "listening test".    You will be able to pick it reliably.... because that's how you picked it in the first place.   It's the basic tenant of the claim  (ie. It's a significant improvement... that can be reliably heard).

 

All I am suggesting is a casual trial of that basic claim to demonstrate it.     ie.  If we listen to some tunes for 60 minutes, and "change the cable" every 5 minutes  (or whatever other combination of listening and changing someone feels is best) ..... can I tell reliably that it was really changed, or just "fake changed".

 

As a second layer.....  we could test if out of the 12 ..... If I could tell which was the improved and which was the not.

 

 

It is just basic "listen to your stereo and see if you can hear a thing or not".    Isn't that what people do all the time.   Isn't this what the "anti measurement crowd" cry out for, in preference to some other measuring stick.

 

People make these claims, and then get upset when people say "oh, that's impossible" .... or "oh, your doing it wrong", and I do genuinely understand their sentiment there.

 

.... but when someone says, "oh, ok.... let's see (I should say, hear)" .... people lose their mind.

 

 

14 minutes ago, frednork said:

there is no measurement thus far to confirm this "big thing" change

 

This issue depends on the specific question you are addressing.... and what you want to be able to say, with what certainty, about cause and effect.

 

In my above example, it is simply "could a difference be determined" (or, assigned to improved/not, correctly) ..... we do not know if the signals are really different (because they have not been quantified) .....  but we have the premise that they ARE different, because there is an audible difference claimed.

 

The test attempts to demonstrate that premise.     If it cannot, then we might start to think that said premise is not as clear as was claimed... but there are still many unknowns.

[frednork]

2 minutes ago, davewantsmoore said:

No.   I am not.

 

We have a claim that "there is something significant" (my paraphrasing).

 

We don't know what "audio measurements" of it look like....   we could try to get them .... but then there are also (sometimes) claims that it is audible no matter whether or not it is measurable.    I dispute that (the something audible could me "un-measurable") .... but there is no need to get tied up on that.

 

Any significantly audible thing .... will pass a "listening test".    You will be able to pick it reliably.... because that's how you picked it in the first place.   It's the basic tenant of the claim  (ie. It's a significant improvement... that can be reliably heard).

 

All I am suggesting is a casual trial of that basic claim to demonstrate it.     ie.  If we listen to some tunes for 60 minutes, and "change the cable" every 5 minutes  (or whatever other combination of listening and changing someone feels is best) ..... can I tell reliably that it was really changed, or just "fake changed".

 

As a second layer.....  we could test if out of the 12 ..... If I could tell which was the improved and which was the not.

 

 

It is just basic "listen to your stereo and see if you can hear a thing or not".    Isn't that what people do all the time.   Isn't this what the "anti measurement crowd" cry out for, in preference to some other measuring stick.

 

People make these claims, and then get upset when people say "oh, that's impossible" .... or "oh, your doing it wrong", and I do genuinely understand their sentiment there.

 

.... but when someone says, "oh, ok.... let's see (I should say, hear)" .... people lose their mind.

 

 

 

This issue depends on the specific question you are addressing.... and what you want to be able to say, with what certainty, about cause and effect.

 

In my above example, it is simply "could a difference be determined" (or, assigned to improved/not, correctly) ..... we do not know if the signals are really different (because they have not been quantified) .....  but we have the premise that they ARE different, because there is an audible difference claimed.

 

The test attempts to demonstrate that premise.     If it cannot, then we might start to think that said premise is not as clear as was claimed... but there are still many unknowns.

what you are describing is a attempt to "hear a small audible difference" with a very lax protocol. Its really not worth attempting unless you are happy to accept that an inability to hear a difference shows nothing at all.  The further you are  away from this https://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1116-3-201502-I!!PDF-E.pdf the less chance your test is of any value. 

 

You seem to be assuming that humans can measure as reliably as your testing equipment. The whole purpose of the standard above is to "convert" notoriously unreliable human subjective impressions into a "reasonably" reliable and repeatable result.

 

If you are using your "super audio analyzer 3000" to measure at the limits of its resolving power, to get an accurate result you need to make sure the machine is setup and calibrated perfectly to extract that last bit of resolution possible.  The standard above also does that for measuring audio differences by humans. If you dont "setup" and "calibrate" you cannot be sure if the "machine" is working properly. 

 

The test is very skewed towards "finding a difference" result. Almost any protocol (no matter how simple) that can be repeated, which shows a difference can be distinguished, is enough to prove that statistically at whatever level of confidence you can achieve, a difference can be heard.

 

However no protocol for this kind of test can show it cannot be distinguished with any statistical confidence and much effort needs to go into showing that it most likely cannot be distinguished. It is totally the wrong kind of test to show things cannot be distinguished but can be used to do that as long as you have appropriate standards, checks and balances in the methodology (which does take some effort which is never done outside academia and sometimes not even there)

 

I expect the vast chasm between what is required to show a difference compared to what is required to show that most likely there is no difference is what confuses people. 

[frednork]

On 29/11/2022 at 6:07 PM, davewantsmoore said:

The computer (or whatever audio device accepts the ethernet connection) is the most important part in this puzzle.

 

If it is susceptible to any sort of ethernet performance (in a way which is allowed to impact the audio) .... then it is doing something wrong... and that should be fixed.

 

 

Not exactly.

 

I would like to argue, that if your ethernet does impact the sound quality..... then that is the problem you should solve.   Not by improving the ethernet, but by improving the device you are connecting to said ethernet.

 

Yes, that;'s what I"m trying to do.

 

23 hours ago, davewantsmoore said:

 

 

It depends on the receiving device... and anything else downstream from it involved in handling the audio.

 

As you vary the noise.... these downstream devices may, or may not, change their performance.... but how and why they change depends on the device.

 

So, there is no universal solution (something that will work to improve, or affect, all audio equipment)  ..... and there is also no universal problem (some things may not be affected by the network).

 

19 hours ago, dbastin said:

It would be better to share "this worked really well for me with this set up" in case it helps someone narrow down the possibilities of what to try in their set up.  And that is the intent of this thread.

 

19 hours ago, dbastin said:

Rather than objectively speculate, please feel free to share any magical endpoint that is immune to any changes to the network it is connected to.   Grimm Audio MU 1 could be one such thing, I gather Chord M Scaler has a similar benefit, and both these rely on code to filter out noise (wherever it may come from).

 

I am more interested in this "perfect" computer (and other connected bits downstream) that is unaffected by ethernet looks like. Could you please describe it further.

 

My listening experience is that even the most expensive "computers" for audio are still affected by switches and cables so very interested in how this could be done.

[davewantsmoore]

20 minutes ago, frednork said:

what you are describing is a attempt to "hear a small audible difference" with a very lax protocol.

No.

We attempting to "hear a significantly audible difference".

 

If the test "passes", then it's fairly certainly really a significant difference.    If the test "fails", then it might not be.

 

20 minutes ago, frednork said:

Its really not worth attempting unless you are happy to accept that an inability to hear a difference shows nothing at all.

Does it really show nothing at all though.

 

"Failing" the test gives us poise to think that the difference may not actually be as "audibly significant" as first claimed.   The reason it does, is that if it were really significantly audible, then it would have easily passed the test.

 

This is of course, only valid for things which are really properly reliably audible.    As soon as we get down into things which are somewhat difficult to reliably hear, then the test will almost certainly show nothing.

 

 

This is why, if I was at the test, I would pay attention to a small subset of the participants, who WERE reliably able to tell the difference.    Perhaps it not something that I can hear ... perhaps its not something I am "trained" to hear.   Whatever, it would be an interesting result.    "Proof" of nothing, but something to consider/investigate/ponder.

 

A claim of "when I do X in my system I can reliably hear it" .... deserves to be tested.    With just a basic protocol:

 

"Do the thing you said"

with a basic amount of blinding / control sample

Is it obvious now?

 

[davewantsmoore]

29 minutes ago, frednork said:

My listening experience is that even the most expensive "computers" for audio are still affected by switches and cables

Did you take my suggested test?

 

All you need to do is keep listening notes from your regular listening sessions.... and not know which cable is connected (your "audio stuff" .... vs a straight through cable from "noisy land") ... you might need an assistant.

 

 

The computers I';m talking about are just basically standard computers.... connected to some sort of good quality audio device (with quite aggressive reclocking, and power isolation from the computer/dirty side using a transformer or batteries).

 

I can literally "dos" attack my audio computer (flood it over the network), and not notice.... or replace the cable and router (and it's PSU) connected to the computer with whatever horrible thing you care to name.

 

 

 

[frednork]

7 minutes ago, davewantsmoore said:

No.

We attempting to "hear a significantly audible difference".

 

If the test "passes", then it's fairly certainly really a significant difference.    If the test "fails", then it might not be.

 

Does it really show nothing at all though.

 

"Failing" the test gives us poise to think that the difference may not actually be as "audibly significant" as first claimed.   The reason it does, is that if it were really significantly audible, then it would have easily passed the test.

 

This is of course, only valid for things which are really properly reliably audible.    As soon as we get down into things which are somewhat difficult to reliably hear, then the test will almost certainly show nothing.

 

 

This is why, if I was at the test, I would pay attention to a small subset of the participants, who WERE reliably able to tell the difference.    Perhaps it not something that I can hear ... perhaps its not something I am "trained" to hear.   Whatever, it would be an interesting result.    "Proof" of nothing, but something to consider/investigate/ponder.

 

A claim of "when I do X in my system I can reliably hear it" .... deserves to be tested.    With just a basic protocol:

 

"Do the thing you said"

with a basic amount of blinding / control sample

Is it obvious now?

 

 

If I squint my eyes and look at your comments from a distance they look to me much more about trying to disprove any difference may exist than doing everything possible to prove they do, as you seem to be suggesting that an unmeasurable difference can be distinguished using a  simple protocol. This is the perspective of someone who does not think differences exist, and as I have said many times, such people should not be involved in the design of such tests.

[frednork]

5 minutes ago, davewantsmoore said:

Did you take my suggested test?

 

All you need to do is keep listening notes from your regular listening sessions.... and not know which cable is connected (your "audio stuff" .... vs a straight through cable from "noisy land") ... you might need an assistant.

 

 

The computers I';m talking about are just basically standard computers.... connected to some sort of good quality audio device (with quite aggressive reclocking, and power isolation from the computer/dirty side using a transformer or batteries).

 

I can literally "dos" attack my audio computer (flood it over the network), and not notice.... or replace the cable and router (and it's PSU) connected to the computer with whatever horrible thing you care to name.

 

 

 

Ok it seems you are saying any reasonable setup will be impervious. Sorry you have lost me there, that does not align with my experience.

[davewantsmoore]

1 hour ago, frednork said:

If I squint my eyes and look at your comments from a distance they look to me much more about trying to disprove any difference may exist

I don't know what would happen.

 

You have to take someones comments at face-value.... and such a test would be my attempt to do so.

 

1 hour ago, frednork said:

as you seem to be suggesting that an unmeasurable difference

We don't know that it is unmeasurable.

 

If we go with the premise that the difference is audible (that is what is claimed) .... then I will assume that the difference will be measurable (somehow, somewhere).    But we don't need to get hung up on that for just an every-day, "let's 'ave a look then".

 

People want listening impressions to be "accepted", and want people to listen rather than measure .... and that's precisely what I'm suggesting.    Just a casual listening session with basic blinding of a control sample.    It's not that hard to setup (especially in this specific example) ..... and I might find the kool-aid delicious.

 

1 hour ago, frednork said:

This is the perspective of someone who does not think differences exist

I would be very surprised... but I don't not know whether a difference exists, in this example.

 

1 hour ago, frednork said:

, and as I have said many times, such people should not be involved in the design of such tests.

What?

How does my "personal perspective", dictate whether or not I should "design a test".   Would it be ok, if someone who did think "differences exist", were to design the same test?

 

The test isn't a "trap".... it's just the most simple, "let's have a listen" .... comparing the "thing" to "the system without the thing" .... and avoiding basic "sighted bias".

 

For said general purpose.... I claim that "the test" is fine.   It is doing nothing more than sitting around listening to a stereo like normal people do when they're trying A and B.    It will be enough to see if something that is thought to be significantly audible, is significantly audible.

 

 

I said before you were over-complicating things.... and now I think I may have made an understatement.

 

[davewantsmoore]

1 hour ago, frednork said:

Ok it seems you are saying any reasonable setup will be impervious

No.  It's quite possible that it will not be.

 

I go to fairly extreme lengths to keep my computer away from my audio devices.    Mains transformers and/or special cables, and/or batteries ... and the audio devices use quite aggressive reclockers.

 

My digital audio player (ethernet* in > spdif out), cost over twice as much as my amplfier(s).....  which many who don't think any of this is important and/or "bits is bits" would find totally crazy.

 

 

* I use mainly wifi, FWIW ... but ethernet cable works just the same.

Edited December 1, 2022 by davewantsmoore

[NPC]

46 minutes ago, davewantsmoore said:

and I might find the kool-aid delicious.

 

Yep.

 

I am beginning to suspect that some streamers perform a level of noise rejection that there is not very much that can be done to shake them. The inquisitive mind is wanting to understand the foundation to the claims being made and it seems some folk want to exclude the very people from testing who have a healthy curiosity and scepticism too. Aren't these the very people that the proponents of expensive switches and cables want to eat their words?

 

If shown - and that means working all the way through the chain to show the suspicion that some streamers are better than others and upstream changes have no (or undetectable) impact on them - it won't be very satisfying from a gloating perspective. When science changes its mind, it wasn't wrong based on what was known before, instead, science learnt something new.

Edited December 1, 2022 by El Tel

[frednork]

2 hours ago, davewantsmoore said:

What?

How does my "personal perspective", dictate whether or not I should "design a test".   Would it be ok, if someone who did think "differences exist", were to design the same test?

 

The test isn't a "trap".... it's just the most simple, "let's have a listen" .... comparing the "thing" to "the system without the thing" .... and avoiding basic "sighted bias".

 

For said general purpose.... I claim that "the test" is fine.   It is doing nothing more than sitting around listening to a stereo like normal people do when they're trying A and B.    It will be enough to see if something that is thought to be significantly audible, is significantly audible.

 

 

I said before you were over-complicating things.... and now I think I may have made an understatement.

Problem is, your assumption is the difference is big. If you were wanting to compare speakers I would say your method is probably ok but would be verified by either picking a difference with different speakers or if not then would say, we dont know if the speakers are different or the test is causing problems.

 

You have said yourself that sometimes even differences that you hear sighted and would expect to hear blinded disappear when under test. My question to you is did they disappear or did you not try hard enough in your testing to ensure the differences were discernable. My gut feel is the latter as you have mentioned many times about not telling people what is different, about misleading people as to what is being tested etc. These are the actions of someone who does not want to find a difference. Look at the standard above and see what it says about what should be told to the participants.. 

 

2 hours ago, davewantsmoore said:

The test isn't a "trap".... it's just the most simple, "let's have a listen" .... comparing the "thing" to "the system without the thing" .... and avoiding basic "sighted bias".

 

But it is a trap, as those that are not familiar with the requirements for the kind of test you suggest would think your suggestion is reasonable, but once you are blinding and testing it is not a casual listen anymore and cant be treated as such. You cant keep saying that sighted listening is fundamentally flawed due to the impact of external factors and then say in the same breath that blinding solves all issues. Subjects knowing they are under test is a huge external factor and can change their answers. This is why expert panels exist, to minimize this issue and to maximise the performance of the participants under test as they become used to testing and have lots of training of what to listen for as well as being comprehensively tested to ensure they are able to discern known similar level changes.  

 

With your own tests or anyone elses, how do you/they know that you, or your system, or your experimental technique are capable of discerning a difference at the level you required. Without specific examples of what you have done, or what you can discern using your method its not possible to even make a guess.

Edited December 1, 2022 by frednork