The thing with audiophile networking equipment... where are the proofs?

5 minutes ago, pkane2001 said:

Isn’t agreeing generally against the rules on the internet? 😁

I see it more simply. When everyone agrees with you, all you have to do is sleep. When they disagree, you're forced to be alert, and that's a good thing.

13 hours ago, jabbr said:

There is no need to buy network gear from a local audio dealer!!!

"Need" is a funny word in audiophiledom, isn't it?  There is really no need to spend more than $2-3K on a stereo that one could buy in an hour or so--see @DuckToller's very good thread on this.  But of course, once the bug bites...we NEED better sound.

I do not stream because every time I have listened to a streamer, I have preferred local files.  Maybe I need to try more expensive streamers or better networking?

4 minutes ago, PeterG said:

I also generally accept scientific proof.  But proof in these cases is very hard to define.  Just for example, a solid state amp is almost certain to have less distortion and noise than a comparable tube amp, but that does not mean the solid state amp sounds better.

Perhaps but it implies that it will not sound worse since any additional distortion and noise cannot be improvements.

6 minutes ago, PeterG said:

IsoAcoustics, Nordost, AudioQuest et al have beautiful charts on their website, but a civilian has no way of evaluating their meaningfulness.  Unlike other areas of inquiry, we do not have a guaranteed set of criteria.

Perfect.  It their objective data (as well as any 3rd party objective data) indicate an objective improvement, one would be justified in auditioning the products to determine if their was any audible effect.  In most cases (which include products from such companies as you list), I have not.

11 minutes ago, PeterG said:

Unlike other areas of inquiry, we do not have a guaranteed set of criteria.

Sure, but that would require the sort of statistically reliable test of the sort that is expensive and, thus, rare.  

13 minutes ago, PeterG said:

More importantly--noise reduction is expensive and complicated--I've added over $10K and multiple boxes and wires.  I could have made an audible upgrade to another component for that price, so I'm looking for more than subliminal improvement as a return on the investment.

Seems a little obsessive to me but, then again, I cannot know what noise problems you face.  I've had a few but they were dealt with simply by replacing the offending devices.

3 minutes ago, PeterG said:

I do not stream because every time I have listened to a streamer, I have preferred local files.  Maybe I need to try more expensive streamers or better networking?

I do not stream because streaming sources have not supported multichannel and, so far, ATMOS streaming has been frustratingly hit-or-miss.  So, I prefer local files, too.

13 minutes ago, Kal Rubinson said:

28 minutes ago, PeterG said:

 

Perhaps but it implies that it will not sound worse since any additional distortion and noise cannot be improvements.

C'mon--something is either better, worse, or equal.  Since a significant group of sophisticated buyers/listeners prefer tubes, we should accept that those people believe SS is worse, even while measuring better.  So we do not have proof that one type of amp is better.

16 minutes ago, Kal Rubinson said:

30 minutes ago, PeterG said:

 

Perfect.  It their objective data (as well as any 3rd party objective data) indicate an objective improvement, one would be justified in auditioning the products to determine if their was any audible effect.  In most cases (which include products from such companies as you list), I have not.

I think we agree on this, and it's the most important point 😊

Just now, PeterG said:

C'mon--something is either better, worse, or equal.  Since a significant group of sophisticated buyers/listeners prefer tubes, we should accept that those people believe SS is worse, even while measuring better.

That is not a justified conclusion since it is based on anecdotal, uncontrolled evidence.  

50 minutes ago, Kal Rubinson said:

That is not a justified conclusion since it is based on anecdotal, uncontrolled evidence.  

Virtually all human beings reach justifiable conclusions every day with anecdotal, uncontrolled evidence.  But also, there's nothing anecdotal about observing the buying behavior of an entire market and concluding that some people prefer one design and others prefer another.  You might assert that one group is flawed or that buying criteria are unknown, but in that case you're asking for a level of proof that is getting kind of silly, and the burden of proof is on you.

28 minutes ago, PeterG said:

Virtually all human beings reach justifiable conclusions every day with anecdotal, uncontrolled evidence.  But also, there's nothing anecdotal about observing the buying behavior of an entire market and concluding that some people prefer one design and others prefer another. 

Certainly.  However, all this tells us is: Chacun à son goût.

31 minutes ago, PeterG said:

You might assert that one group is flawed or that buying criteria are unknown, but in that case you're asking for a level of proof that is getting kind of silly, and the burden of proof is on you.

Not expecting any proof (I thought that was clear). 

However, your statement was "Since a significant group of sophisticated buyers/listeners prefer tubes, we should accept that those people believe SS is worse, even while measuring better."  One can also say "Since a significant group of sophisticated buyers/listeners prefer SS, we should accept that those people believe tube are worse and in addition to just measuring worse."  Back to square one.

7 hours ago, PeterG said:

"Need" is a funny word in audiophiledom, isn't it?  There is really no need to spend more than $2-3K on a stereo that one could buy in an hour or so--see @DuckToller's very good thread on this.  But of course, once the bug bites...we NEED better sound.

 

I do not stream because every time I have listened to a streamer, I have preferred local files.  Maybe I need to try more expensive streamers or better networking?

To be very clear. and I've spent considerable amount of time looking into this as well as listening, the quality of my equipment, lets say my 100Gbe Mellanox switch, QSFP modules, cables ... essentially my entire network ... is very simply unmatched by any "audiophile" network equipment in existence. Its not even close. I do grant that endpoints and equipment that sits on the audio rack should be engineered to the requirements of audio.

Now since you don't stream, you are commenting on an aspect for which you admit that you don't have direct knowledge. I haven't "listened" to every piece of network equipment but I can say with a high level of confidence that the single most glaring **reduction** in SQ from a network is by transmission of common-mode noise as well as ground loops (similar things) via Ethernet cables. Once you go above the very cheap network equipment that is in the common home network, and move to 10Gbe which at this point in time is almost legacy, then things like network jitter is irrelevant. 

Regarding "streaming", the SQ from HQPlayer streamed to a low power NAA (which sits in the audio shelf) is nothing short of amazing. The ability of the modern network to definitively isolate the high noise/high powered workstation needed to process HQPlayer at its highest level from the low noise audio area is also nothing short of amazing. I say this from both a technical as well as vast listening perspective.

On 10/24/2022 at 9:53 PM, sine said:

I see now that my choice of words (educated opinion vs. opinion) is problematic and opens up a bottomless can of philosophical considerations, which divert from the intention of my original question.

I agree that the perception of audio can't be solely meassured by engineering means, hence there is the field of psychoacoustics. Nevertheless psychological phenomena related to the listening experience could be scientifically measured, and I'd wish more effort would be made to do so (and to publish the findings), so we can get a grasp of the magnitude these effect(s) can have on us (e.g. by ABX tests [1]).

 

Thank you @Savolax for the reference to the conversation about the ethernet cards [2] - that's the kind of assessment I'd fancy as a starting point.

 

On that page I also came across a network switch test by the author [3]. It amazed me that although one supposedly importent parameter (jitter) technically got worse after attaching an external clock source (which is supposed to be an enhancement), it was claimed that the overall listening experience of the owners of such setups was elevated. Now I've factual reason to belief that jitter may be overrated.

 

Although it sounds silly but Wikipedia's list of cognitive biases [4] seems to be a treasure trove of potential enhancement ideas for the audiophile listening experience ;) and I have no problem with that - just because I have factual knowledge about what a rainbow is doesn't deminish the magic seeing one, or does it?

 

Wouldn't it be nice if we as a community could resort to a set of testing protocols to account for all these factors, and to enable us to compare and peer review our findings? After all doesn't we all seek the most bang for our bucks? I think dividing the community in 'objectivists' and 'subjectivists' is an oversimplification of the matter and misses the point, since psychoacoustics [5] transcends the domains of acoustics, biology, electronic engineering, psychology, physics, physiology, and computer science.

 

 

[1] https://en.wikipedia.org/wiki/ABX_test
[2] https://audiophilestyle.com/forums/topic/64138-best-ethernet-cards-for-streaming/page/7/#comments
[3] https://www.open-end-music.com/forum/privatforen/thomas-michael-rudolph-tmr/651284-messungen-von-ethernet-infrastruktur-switches-nur-lesen/page3#post659024
[4] https://en.wikipedia.org/wiki/List_of_cognitive_biases
[5] https://en.wikipedia.org/wiki/Psychoacoustics

One: all of that is very expensive and very difficult to actually test properly. That's why it's almost never done for audio.

Two: it probably won't actually make much of a difference. People who "believe in their ears" will continue to do so, even if testing and evidence say they are wrong. Those who tend to accept objective testing will continue to accept it vs. sighted listening. 

Thus the endless arguments about it on audio forums, even when such testing is occasionally done.

1 hour ago, firedog said:

One: all of that is very expensive and very difficult to actually test properly. That's why it's almost never done for audio.

Two: it probably won't actually make much of a difference. People who "believe in their ears" will continue to do so, even if testing and evidence say they are wrong. Those who tend to accept objective testing will continue to accept it vs. sighted listening. 

Thus the endless arguments about it on audio forums, even when such testing is occasionally done.

The thing with pro-grade network equipment ie 10Gbe ethernet and faster, is that **all** of it is required to be tested and not pass network jitter, so you can hear a difference, and know that its been compliance tested in a very stringent fashion.

10 hours ago, firedog said:

But you are at the objective forum, which is the one are of the site where the discussion isn't supposed to be about such subjective things. It's the area for discussing objectively, without getting into the endless subective arguments.  So don't try to move the discusion there. You have the other 99% of the site for that. 

It wouldn't cross my mind to move something or someone. Firstly, I don't need it, and secondly, I'm too lazy. And I would certainly not tell someone what to do or not to do. As they say, don't tell me what to do and I won't tell you where to go 😎

Network equipment designed for audio use is expected to not emit acoustic noise.

General network gears (or computer monitors, PC, or other equipment) simply does not care about it.

Some network gears have annoyingly loud cooling fans, another one emit audible acoustic noise from the inductor or MLCC of its switching power circuits that is clearly heard by ears. The other emits ultrasonic acoustic noise: those are not heard from ears and no problem for audio listening but it contaminates high-res sound recording so it is important those who records high-res sound using microphones.

Those problems are all measurable, I had uploaded high-res recorded sound files of my worst sounding network gear on this forum but cannot found now.

2 hours ago, yamamoto2002 said:

Network equipment designed for audio use is expected to not emit acoustic noise.

General network gears (or computer monitors, PC, or other equipment) simply does not care about it.

 

Some network gears have annoyingly loud cooling fans, another one emit audible acoustic noise from the inductor or MLCC of its switching power circuits that is clearly heard by ears. The other emits ultrasonic acoustic noise: those are not heard from ears and no problem for audio listening but it contaminates high-res sound recording so it is important those who records high-res sound using microphones.

 

Those problems are all measurable, I had uploaded high-res recorded sound files of my worst sounding network gear on this forum but cannot found now.

Acoustic noise is a different discussion, of course. Some of the audiophile network switches and cards run much hotter than their non-audiophile, off-the-shelf versions. I don't know if the expectation of no acoustic noise is what some of these devices are really designed for: as I understand it, most are designed with the idea to reduce electrical/electronic noise, not acoustic. That said, I've yet to run into any non-audiophile network switch for home use that has fans built in or makes any audible coil whine. Data centers, of course may have lots of acoustically noisy switches/routers.

My audiophile DAC, on the other hand, developed a significant and audible hum coming from the power supply when plugged in to an audiophile power regenerator. I assume it's caused by a DC offset, but my solution was to scrap the regenerator - problem solved, and not by using audiophile cleanser devices or other tweaks ;)

14 hours ago, Archimago said:

 

You used the word "better" twice here but what is being judged is very different.

 

An SS amp "measuring better" refers to an empirically derived judgment based on things like distortion, or lower noise compared to a ideal. Lower distortion, lower noise, generally being more ideal and therefore "better" - higher fidelity if you will.

 

As for "proof" of "one type of amp is better" based on the the opinions of "sophisticated buyers/listeners" (what is the definition of a "sophisticated" consumer in this context?) is of course subjective judgement based on whatever those listeners deem to be good, pleasant or preferable. Many variables have been unaccounted for - like what the rest of their gear consists of, or the noise floor in their room, or even what sentimental reasons some might have to enjoy the glow of the tubes.

 

I believe that for high fidelity reproduction, SS amps are better and would use that when evaluating high quality DACs in a review. But as a subjective being enjoying some Diana Krall over a nice glass of red wine with my wife, there's nothing wrong with thinking the "warmth" of a tube amp in that context is also better. 😉

 

I don't think there's any "audiophile police" forcing a person to choose tube vs. SS. But for high fidelity reproduction with lowest distortion, generally the SS amp is better so long as you're staying within the power limits.

I think we are in complete agreement, or at least we would be if we were trading thoughts in person.  I was using the word "better" in an attempt to say that there is no universal better in this context, just as there is no "better" between salmon and steak.  I agree that SS are better on paper.  But I do not accept that better on paper does anything more than give us an indication of some aspects of how the amps will sound.  In practice, which I take to be the real definition of better, equally discerning sophisticated listeners are split between the platforms.

Your two examples of listening are understandable, but kind of on the guardrails of typical audiophile listening.  On the SS DAC reviews, you are not listening for pleasure, you are testing, notebook in hand.  On the tube date, you are barely listening at all, you are trying to seduce a beautiful woman.  I hope you are successful with both!  But for me the real test is a single person, alone and relaxed, being moved by life-like sound

1 hour ago, semente said:

 

I disagree. There is a universal definition for better performance: accurate reproduction of the recorded signal.

 

Better-sounding, on the other hand, depends on the listener's preference. It's subjective and thus not universal.

This is very true! The SS vs tube amp debate is drastically different than when we are discussing network gear. Many expert amp designers understand that various odd vs even harmonics can provide enjoying sonic coloration. Ultimately this is about listening pleasure. 

Network equipment do not have a role in tuning SQ. Any sound introduced to the system by a piece of network gear is a fault.

Just a friendly reminder to some that this post is in the Objective-Fi forum. 
 

🙂

On 10/26/2022 at 5:48 PM, yamamoto2002 said:

Those problems are all measurable, I had uploaded high-res recorded sound files of my worst sounding network gear on this forum but cannot found now.

Found it. It is about annoying acoustic noise from switching mode power supply circuit.

On 10/26/2022 at 6:07 PM, The Computer Audiophile said:

Just a friendly reminder to some that this post is in the Objective-Fi forum. 
 

🙂

Thank-you!  I had forgotten this forum existed and I could not figure out how I had stumbled into some strange parallel universe. (Seriously, I am not kidding)

Good luck, gentlemen.  I bid you adieu 😊