advertisement


Audiophile Network Switches for Streaming ... really ?

A: There's not any difference in sound using a CAT6 or a CAT8 cable
B: But I bought a 250 clam USB-cable and heard an improvement
 
Discopants said:
I can't help with the why and wherefore :) I'd just like to say that I am also mad and happily have spent a fair bit of cash on the placebo effect of streaming optimisation.

I've not read the whole thread but this question comes up on fora very regularly and there is always a hardcore of network technicians who exclaim its impossible because of TCP/IP protocols and they have to stop the snake oil. A hardcore of folks like me who took the blue pill and are convinced they can hear significant differences and a few folks who have experimented with an open mind and heard nothing.
Click to expand...

Nice post.

The trouble with IT guys (spolier: I am one) is that we quite understandably first look at switches in the audio world through exactly the same lens we use for our day job (just as I did 12 months ago). This inevitably leads to an argument which I shall split into two parts:
  1. Bits is bits
  2. Therefore a (digital) network switch cannot possibly have any impact on sound quality
The first part of this is 100% spot on. No switch should be claiming to massage bits so they sound better, and this includes a whole shoal of red herrings about clock accuracy in the ethernet domain.

The second part is complete tosh. A switch can demonstrably impact sound quality but this is due entirely to its action in the analogue (RFI/EMI) domain; a side effect of the design of most ethernet circuitry which audiophiles and audio companies have exploited. The switch needs to be installed just before the streamer not just at random in the digital chain. The only differentiator amongst network switches deployed for such audiophile purposes - from budget Zyxels and the like to some seriously expensive kit - is the amount of noise they ultimately add to or mitigate in your system.

Variables of course include the noise sensitivity of your kit and the senstivity of your ears, but I and others have been clearly able to differentiate switches in a £100k+ system which one would like to think is probably designed to be pretty noise-resistant in the first place.
 
TheFlash said:
Variables of course include the noise sensitivity of your kit and the senstivity of your ears, but I and others have been clearly able to differentiate switches in a £100k+ system which one would like to think is probably designed to be pretty noise-resistant in the first place.
Click to expand...
You may have heard the difference or you may have imagined these differences. We all do the latter from time to time. Now the question is how reliably these findings could be repeated in a proper blind test.
 
That's silly. How many non-audiophile switches are there between the download source (?) and your socket? A hundred? Possibly much more... I can't imagine the data loss.
And there might even be radio transmission into the chain. The horror! Data lost in the ether, oh dear.
Let's be serious for a minute.
 
Octavian said:
You may have heard the difference or you may have imagined these differences. We all do the latter from time to time. Now the question is how reliably these findings could be repeated in a proper blind test.
Click to expand...
Ah yes, the blind test which relies on unreliable aural memory. There may be some merit if a large enough sample with methodology that allows for both comparison of short extracts of sound and long term listening to music, but who on earth is going to fund that for such a niche market as hifi. It ain’t going to happen, which is probably why hard core objectivists like to use it as a rejoinder and crusading weapon.

Not really comparable but I’m always reminded of that Harman test of speakers which used a shuufleometer to move speakers in and out without the listener knowing what they were listening to. What a shame that they didn’t consider that different speaker types need different setups with regard to room and listener positioning before claiming that their owne speaker was very good and a rival type of speaker was very bad! Even worthy attempts at blind tests are fraught with problems.

Consider the topic under discussion, apart from the large number of participants required there would also need to be agreement on the partnering equipment. As I’m sure we are all aware some equipment can be more critical for revealing noise in the chain than others. The real question is why bother insisting on a test that isn’t going to happen in any meaningful form - unless of course you are just trying to win an argument or pusue an agenda!
 
camverton said:
Ah yes, the blind test which relies on unreliable aural memory. There may be some merit if a large enough sample with methodology that allows for both comparison of short extracts of sound and long term listening to music, but who on earth is going to fund that for such a niche market as hifi. It ain’t going to happen, which is probably why hard core objectivists like to use it as a rejoinder and crusading weapon.

Not really comparable but I’m always reminded of that Harman test of speakers which used a shuufleometer to move speakers in and out without the listener knowing what they were listening to. What a shame that they didn’t consider that different speaker types need different setups with regard to room and listener positioning before claiming that their owne speaker was very good and a rival type of speaker was very bad! Even worthy attempts at blind tests are fraught with problems.

Consider the topic under discussion, apart from the large number of participants required there would also need to be agreement on the partnering equipment. As I’m sure we are all aware some equipment can be more critical for revealing noise in the chain than others. The real question is why bother insisting on a test that isn’t going to happen in any meaningful form - unless of course you are just trying to win an argument or pusue an agenda!
Click to expand...

It would be a poorly-designed double-blind test that did rely on aural memory, but you’re right, and as I posted when this thread was still in its infancy, a double-blind test capable of ending this never-ending argument is not going to happen.

That said, accusing those who bring it up as having an agenda is very unfair! I’m not going to repost the comments and helpful posts on expectation/confirmation/dominant participant bias from earlier, but pretending they don’t exist or is used only to block a path to the truth does not further the discussion.
 
TheFlash said:
. A switch can demonstrably impact sound quality but this is due entirely to its action in the analogue (RFI/EMI) domain; a side effect of the design of most ethernet circuitry which audiophiles and audio companies have exploited. The switch needs to be installed just before the streamer not just at random in the digital chain. The only differentiator amongst network switches deployed for such audiophile purposes - from budget Zyxels and the like to some seriously expensive kit - is the amount of noise they ultimately add to or mitigate in your system.

Variables of course include the noise sensitivity of your kit and the senstivity of your ears, but I and others have been clearly able to differentiate switches in a £100k+ system which one would like to think is probably designed to be pretty noise-resistant in the first place.
Click to expand...
You can substitute "your neighbours' toaster" for network switch and entirely retain the almost immeasurably small amount of truth in this post.
 
camverton said:
Ah yes, the blind test which relies on unreliable aural memory. There may be some merit if a large enough sample with methodology that allows for both comparison of short extracts of sound and long term listening to music, but who on earth is going to fund that for such a niche market as hifi. It ain’t going to happen, which is probably why hard core objectivists like to use it as a rejoinder and crusading weapon.
Click to expand...
It seems to follow from this then that there is no reason to take hobbyist testimonials into serious consideration. Pity indeed!
 
Why do people trust their ears and brains with this. It's all an illusion. It couldn't possibly be my mood, day I've had - literally anything in this big sack of chemicals but people trust them like it's a golden reference - then start claiming they can hear RFI
 
Warszawa said:
Page 89 reminder that all you need to do to test this theory with most streamers is to have someone pull the network cable during playback.
Click to expand...
Excellent! About 2 minutes or more will still be played with the cable pulled out. The magic of cache. The router does nothing of its own, and neither do the cables. Real time digital doesn’t exist, does it.
 
Addressing no specific position, the latest "audiophile network switch" "thing" seems to me to be another example of what I have observed over 40 or so years. That the audio industry frequently over-promises its new ideas and under-delivers.

I observe that the industry tries to make the buying population dissatisfied with what they have, so they spend more money. This is not uniquely a technique used by marketing for audio.

The worst end of the "new idea" spectrum seems to be the "not understood by science and engineering" tweak justified by spurious, tenuous or exaggerated argument. It looks to me like most of those tweaks promoted during the time I have been interested in audio have failed to gain credible-level justification despite their accumulating years.

Perhaps I have become too cynical as a result. But I do recall the past. And that tells me to not accept without question the assertions of those who market the latest audio ideas, and to be wary of enthusiasm from some who enjoy having the latest products. I do not accept the "just try it" argument. Experience tells me to adopt the "give me a properly argued reason before I will go to the trouble" position.

On the topic of the thread title I have heard nothing that could be described as "wrong" or in need of some "cure" that an audiophile switch might provide. That is what I hear based on my own audio references.

It's a hobby and the above is just my way. I am not criticizing other approaches such as buying the latest audio ideas - audiophile switches included. The only thing I think should be avoided is evangelism that actually or implicitly has a go at others' different but perfectly valid ways of enjoying the hobby. And BTW, "sauce for the goose is sauce for the gander", of course.
 
You must log in or register to reply here.


advertisement


Back
Top