May I remind EVERYONE posting in this thread to follow forum guidelines as can be found here and to show fellow members the respect they deserve and remember we are all in this together.
Not everyone is going to agree or share the same viewpoints but we need to be more considerate in our posting style.
Thank you for your patience and understanding.
1 Like
This thread is a great read, goto take my hat off to @Graeme_Finlayson for his outstanding contributions, and will add I’m firmly with him on this.
5 Likes
Thank you for your kind words.
2 Likes
I enjoy reading this, as I understand the empirical point of view it is somehow accepted by all. I tend to be influenced by theory but that is my bias as sometimes I do not trust myself in the realm of perception so logical argument has the upper hand. What I do not understand is why some of us who are interested in this, cannot get together with all kinds of gear and have all the tests, then post them online for the remote part of the earth (if this was done forgive my ignorance).
Do not get me wrong, I do not want to offend anyone, it is my just ignorance to blame, but I read and I truly and genuinely agree with both/all sides.
Maybe some manufacturers of some audiophile switches and exotic ethernet cables could chime in here? Some claim that re-clocking the digital data stream in their own switches greatly improving the SQ.
1 Like
Really? There is no almost no timing (audio relevant) involved at this stage. As a matter of fact data packets on a 10 MBit/s ethernet line travel way faster than any DAC can process these to audio. Packets are being buffered and stored at various stages (switches, computers such as streamers) until processed. No network component could contribute to audio quality beyond the fact their capacity is a fascinating overkill to what is required.
DO NOT MIX UP ANALOG AUDIO AND DIGITAL DATA. JUST DONT.
2 Likes
Cut & paste from a audiophile switch maker’s web site:
Increase Clocking Precision and Stability
Using the same 3ppb 25MHz OCXO oscillator as used in the Statement, individually powered by its own linear power supply and connected directly to the network switch chip, avoiding precision losses from using external master clocks.
The PhoenixNET is the realization of Innuos’ philosophy of simplicity and signal purity applied to the network switch. Having started with improvements to the Ethernet ports’ clock on our flagship Statement, Innuos has now taken the concept to the next level with a completely new network switch design that focuses exclusively on audio use.
PhoenixUSB
Reclocker
1 Like
Only 3.000 EURO 
Must be great 
2 Likes
I’ll tell you about my experience.
I have a DIY “audiophile” switch. What is this? It’s a €20 standard Netgear switch in which a friend who is an electronics technician replaced the original capacitors that power the synchronization clocks by better ones. This upgrade cost me €35 for the components and the work.
This friend explained to me that the cheap common switches are perfect for office appliances, but not for a continuous streaming of audio and video. By replacing their capacitors, the clocks are better powered, and perform better for a continuous audio data streaming.
The technician also removed some components for superfluous, and disabled features of the switch that are not useful for streaming on the LAN in order to reduce to minimum the noise that the device produces.
When I compared the upgraded Netgear switch to the same Netgear model with its original capacitors, there was a small difference. A slight metallic timbre was removed from the sound.
Recently, I had the occasion to compare this DIY switch to an EtherRegen, and there was no tangible difference at all between them. The sound was quite the same.
However, when I connected my external 10 mHz reference Master Clock to the EtherRegen, and synchronized it this way with other devices of the setup, there was a nice improvement of the sound quality.
I will concede being utterly wrong and give EUR 500 to an animal welfare org of your choice if you prove to tell any “audiophile” network equipment (switches, including cables) from basic Amazon grade gear with any statistical relevance. Blind test, of course.
1 Like
More voodoo magic and snake-oil marketing, I’m afraid.
5 Likes
lol, I just quote the manufacturer’s ads to show that that there are many strong believers in such switches and willing to pay a vast amount of dollars for those, even quite a few in thread.
2 Likes
More utter tosh from Innuos:
Yeah, those darn LEDs. I better go tell John Siau (Benchmark’s Engineering Director) that the LEDs on their gear are compromising their superlative measured performance…
Facts now. Have you ever been able to identify audiophile network gear without having any idea what’s hooked up at the moment? If you can tell an audiophile switch from an 20 bucks model 9/10 times - blindfolded - then we are talking. Else it’s just bias. Easy task …
1 Like
This is the issue in this industry. Fanciful claims with no basis in reality, never mind in science or engineering being used to peddle nonsensical equipment upgrades to solve non-existent problems.
3 Likes
It doesn’t even show that, does it? No idea if the manufacturer (or their ad agency) believes or doesn’t in what they’re saying. But they have to say something to have content in the ad. So what are they going to say? “This switch has no real advantages over an off-the-shelf TP-Link box, but it costs a lot more, so you might feel better having it instead.” Doubt that would work.
1 Like
I don’t have this problem of bias.
I work regularly on the production of soundtracks, and I decide thousands of times a day what small adjustment sounds better.
1 Like
What does this even mean? Networks transfer data and, for the most part, do it with 100% accuracy. How does ‘better’ power to the clock possibly make any difference? 101% accurate? Better than accurate?
As has been stated in numerous recent threads, data transfer over a network is a solved problem.
3 Likes
Plain wrong. Everybody is biased. Proven over and over again.
Again: Could you do this blindfolded?
1 Like
It’s not a problem of accuracy. The DAT remains 100% accurate. But poorly powered synchronization clocks create very short latencies in the regularity in which the packages are continuously streamed. The DATA remains bit-perfect of course.