The Ethernet cables, Switches and Network related sound thread. Share your listening experience only.

[Mark200]

I can see a few ways a fiber optic connection can provide a benefit:

1. EMI/RFI cannot enter the cable itself
2. Provides galvanic isolation between source and end point, eliminating any ground plane noise
3. Reconstructs the signal at the terminal end to reduce jitter
4. Lowers system speed from 1G to 100MBPS, which allows end point gear to work more quietly (although some gear cannot operate at 100MBPS)

Note that ADOT is directly affiliated with Melco, and the recommended use is to have an ADOT SFP in a Melco switch, into a fiber optic cable, with an ADOT FMC and SFP at the audio system end. If a fiber cable could truly shed all digital noise, why would you need an expensive Melco S100 feeding the fiber system? If no noise can come through from the source, upstream gear wouldn't matter, you would be able to use a cheap D-Link switch with SMPS feeding a second ADOT FMC.

I don't understand how fiber lowers the connection speed. Fiber Channel (not talking about optical audio Toslink) is generally capable of much higher speeds than copper Ethernet cable.

Also, if the device at the other end of a 1 Gb router (music streaming device, TV, etc) only supports a maximum speed of 100 Mb (as most streaming devices do), then the 1 Gb router automatically makes the connection at 100 Mb speed even if other connections (to your PC for example) are at 1 Gb.

 

[audiobomber]

I don't understand how fiber lowers the connection speed. Fiber Channel (not talking about optical audio Toslink) is generally capable of much higher speeds than copper Ethernet cable.

Fiber doesn't affect the connection speed, but the ADOT MC-01 kit can decrease the downstream speed to 100MB.

Also, if the device at the other end of a 1 Gb router (music streaming device, TV, etc) only supports a maximum speed of 100 Mb (as most streaming devices do), then the 1 Gb router automatically makes the connection at 100 Mb speed even if other connections (to your PC for example) are at 1 Gb.

Apparently not all audio devices will work at the lower speed:

"Externally, the ADOT MC01 is a generic looking small black box, one side with both SFP and Ethernet plugs, plus LED indicators of status, and the other with a 5v input and 4 small dip switches, which amongst other things, can be used to drop the output from 1GB, down to 100mps in the quest of improved sound quality."
"I paired the ADOT with the Melco S100 network switch and did attempt to feed it with the 100mps output, selected by flicking on switches 3 & 4, followed by power down and back up. Unfortunately, neither my Moon 280D or indeed my Chord 2Go, would work in this setting. A chat with the distributor confirmed that, indeed, the 100mps output is not compatible with certain brands, Linn being one, in particular, that is known to them. They were surprised, but not shocked at the results with my Moon and Chord streamers."
https://www.hifipig.com/adot-mc01-fibre-network-kit/

 

[Mark200]

Fiber doesn't affect the connection speed, but the ADOT MC-01 kit can decrease the downstream speed to 100MB.

Apparently not all audio devices will work at the lower speed:

"Externally, the ADOT MC01 is a generic looking small black box, one side with both SFP and Ethernet plugs, plus LED indicators of status, and the other with a 5v input and 4 small dip switches, which amongst other things, can be used to drop the output from 1GB, down to 100mps in the quest of improved sound quality."
"I paired the ADOT with the Melco S100 network switch and did attempt to feed it with the 100mps output, selected by flicking on switches 3 & 4, followed by power down and back up. Unfortunately, neither my Moon 280D or indeed my Chord 2Go, would work in this setting. A chat with the distributor confirmed that, indeed, the 100mps output is not compatible with certain brands, Linn being one, in particular, that is known to them. They were surprised, but not shocked at the results with my Moon and Chord streamers."
https://www.hifipig.com/adot-mc01-fibre-network-kit/

Why are you trying to convert fiber to Ethernet to feed your streaming devices from an Internet connection? Your ISP router will convert that for you.

If you have a gigabit (1 Gb) router and a 1 Gb streaming device, you can usually lower the speed to 100 Mb by just connecting 4 of the 8 wires (or 2 of the 4 twisted pairs) of the cable from the RJ-45 plugs (I have done this because I make and terminate my own Ethernet cables). If your streaming device is auto-sensing, then it will automatically slow it down when it senses that only 4 of the wires are connected. Obviously, you have to only connect the correct 4 wires (usually the middle 4).

I suppose that it is possible that some streaming devices would not work at 100 Mb, but I doubt that, and the problem is probably something related to using the ADOT MC01, and not a normal 100 Mb connection coming from a router that automatically supports 1 Gb or 100 Mb.

But the whole idea of slowing an Internet connection to a streaming device down to 100 Mb to improve the quality of the signal is not credible unless the Ethernet cable has a very long run. The reason why most streaming devices are limited to 100 Mb is that 100 Mb is all that is needed, and prevents one user from hogging all the bandwidth of the home network when multiple people are using it for computers, streaming, etc.

 

[iFi audio]

The combo which is my priority at the moment is: Bartok to Benchmark AHB2 to Utopia 2022. The best way to characterize the change is that it sounds dryer and less rich. It's an overall less "excitable" kind of sound and also less glare with upper-midrange spikes in the music. There's less lingering sparkle in the treble. Possibly less bass, too? It's drinking your morning coffee with less cream and sugar. The soundstage sounds less filled-in. It may just need some getting used to, but even after "getting used to it", the jury's out whether I will enjoy it.

That's interesting. Usually the sound with quality network switches gets more rich, weightier and smoother, but in your post it's the other way around it seems. Thanks for your feedback nonetheless

 

[ChJL]

I'm using a streamer with integrated storage medium and I play music exclusively from said hard drive and not from a streaming service.

The N100 has to be connected to the internet/router in order to be operated via app on a smart device... (One could play music by using the control buttons on the streamer itself and then I could unplug the ethernet cable which I don't because of inconvenience)

Why would an audiophile switch improve the sound when I don't use streaming services but rather play music from the internal disk?
I mean does "noise" creep in just because there is a connection to the router?
thanks!

 

[dougms3]

Why are you trying to convert fiber to Ethernet to feed your streaming devices from an Internet connection? Your ISP router will convert that for you.

If you have a gigabit (1 Gb) router and a 1 Gb streaming device, you can usually lower the speed to 100 Mb by just connecting 4 of the 8 wires (or 2 of the 4 twisted pairs) of the cable from the RJ-45 plugs (I have done this because I make and terminate my own Ethernet cables). If your streaming device is auto-sensing, then it will automatically slow it down when it senses that only 4 of the wires are connected. Obviously, you have to only connect the correct 4 wires (usually the middle 4).

I suppose that it is possible that some streaming devices would not work at 100 Mb, but I doubt that, and the problem is probably something related to using the ADOT MC01, and not a normal 100 Mb connection coming from a router that automatically supports 1 Gb or 100 Mb.

But the whole idea of slowing an Internet connection to a streaming device down to 100 Mb to improve the quality of the signal is not credible unless the Ethernet cable has a very long run. The reason why most streaming devices are limited to 100 Mb is that 100 Mb is all that is needed, and prevents one user from hogging all the bandwidth of the home network when multiple people are using it for computers, streaming, etc.

Your isp will convert fiber for you by the cheapest method possible because you're not paying for that. Just like they will provide you with the cheapest crappiest router if you decide to get one from them.

 

[dougms3]

I'm using a streamer with integrated storage medium and I play music exclusively from said hard drive and not from a streaming service.

The N100 has to be connected to the internet/router in order to be operated via app on a smart device... (One could play music by using the control buttons on the streamer itself and then I could unplug the ethernet cable which I don't because of inconvenience)

Why would an audiophile switch improve the sound when I don't use streaming services but rather play music from the internal disk?
I mean does "noise" creep in just because there is a connection to the router?
thanks!

I wondered about this myself.

The answer is probably dependent on each individual and the environment.

When i upgraded the Ethernet cable from the verizon fmc to my router, i had zero expectations. I play most music from files on a non nvme ssd drive attached in my pc.

The result was better picture quality and audio quality from streaming services but also the sound quality playing from files improved as well. Which i don't fully understand why, but i was also thinking that there was some noise creeping in somewhere.

Unfortunately I'm restricted to using wifi to connect my pc to the network, not ideal but i have no choice.

 

[kingoftown1]

Similar situation here -- I'm running a fairly optimized audio pc playing files from local storage, wifi disabled, connected with a furutech lan-8 ncf & baaske mi-1005. Both the furutech cable and baaske gave sq improvements, as did adding an unused hdplex lps to the connected mesh satellite. I've been looking at switches and filters as possible next steps -- the NA Eno/Muon (Pro?), Ediscreation or LHY switches. IDK how much to invest before I hit diminishing returns, and I don't want a nest of spaghetti as the result.

 

[llamaluv]

Apparently not all audio devices will work at the lower speed:

"[...] Unfortunately, neither my Moon 280D or indeed my Chord 2Go, would work in this setting. A chat with the distributor confirmed that, indeed, the 100mps output is not compatible with certain brands, Linn being one, in particular, that is known to them. They were surprised, but not shocked at the results with my Moon and Chord streamers."

LOL, I came across that review while reading up on the ADOT kit after @Roybenz mentioned the ADOT, and came across that caveat, which really stuck in my throat.

I'm glad I looked into the ADOT if only because I never looked into the whole world of ethernet-to-fiber before, and I'm pretty intrigued by taking this approach for our purposes.

But as for the ADOT offering specifically, the hardware appears to be a bog standard rebadge of some pre-existing enterprise IT offering (maybe from 10GTek? not sure, as all this is new to me). Even the highly touted feature of being able to configure (albeit unreliably, apparently!) the connection to 100Mbps is no big shakes, and seems to exist on this off-the-shelf ~$25 unit (Amazon link; PDF docs showing 100Mbp DIP switch setting).

As far as I can tell, the only value the ADOT "MC01" kit brings is pre-selecting for the customer a correct, compatible combination of MC units and cable, which to be fair is something of a PITA to navigate. But at a something like 500% markup, that's pretty eye-roll-worthy. The "MC03" kit adds a legit LPS into the mix, at least, but yea, overall, pretty gross. IMO.

Having said that, what I've done is order a couple 10GTek media converter units from Amazon, and a short $10 fiber cable for a grand total of $60. And if the results seem promising, maybe later I'll go for a decent LPS after (maybe a Farad Super3 5V with dual connectors, maybe something cheaper). If I ended up choosing the right parts, I'll share with the thread what I got in case, like me, all this is new to you, too

 

[Mark200]

Your isp will convert fiber for you by the cheapest method possible because you're not paying for that. Just like they will provide you with the cheapest crappiest router if you decide to get one from them.

Most ISP's that provide fiber to inside the home require the use of one of their routers (made by some third party company) that take the fiber input (connected to the Internet) and provides Ethernet output ports for use within the home. There is nothing crappy about them, and they rarely make an undetected mistake, partly because the TCP/IP protocol. TCP/IP transmits data in packets, and there are check-sums and other safeguards to make sure the data has not been corrupted from the source that sent the data, and also to make sure that any missing packets are retransmitted. Virtually the entire computer and Internet infrastructure of the world works with TCP/IP and rarely are there any undetected errors that are not automatically re-transmitted.

If ISP's did allow the errors you claim with "cheap" gear, then no one could trust any financial transaction or account balance, the the ISP would be liable for astronomical liability for bad data that went undetected.

Most people don't realize how many intermediate routers/switches a TCP/IP packet goes through to get from the source to your home. For example, I did a TCP/IP trace on my home computer and it took 14 hops (going through 14 routers/switches) to get from Tidal.com to my home.

The whole idea that computer networks use cheap components, and that streaming hi-res audio needs something superior, is absurd. Computers have zero tolerance for undetected errors and computers would frequently crash if that were not the case.

 

[Mark200]

I'm using a streamer with integrated storage medium and I play music exclusively from said hard drive and not from a streaming service.

The N100 has to be connected to the internet/router in order to be operated via app on a smart device... (One could play music by using the control buttons on the streamer itself and then I could unplug the ethernet cable which I don't because of inconvenience)

Why would an audiophile switch improve the sound when I don't use streaming services but rather play music from the internal disk?
I mean does "noise" creep in just because there is a connection to the router?
thanks!

An audiophile switch will not help you. Even if you were streaming music from the Internet via Ethernet, an audiophile switch will not help, as they are almost always scam devices.

 

[Mark200]

That's interesting. Usually the sound with quality network switches gets more rich, weightier and smoother, but in your post it's the other way around it seems. Thanks for your feedback nonetheless

No, "quality" network switches do not result in "more rich, weightier and smoother" music transmitted over TCP/IP over Ethernet. I am now ashamed that I own 2 pieces of iFi equipment, but I will never buy any more again from your brand.

 

[dougms3]

Most ISP's that provide fiber to inside the home require the use of one of their routers (made by some third party company) that take the fiber input (connected to the Internet) and provides Ethernet output ports for use within the home. There is nothing crappy about them, and they rarely make an undetected mistake, partly because the TCP/IP protocol. TCP/IP transmits data in packets, and there are check-sums and other safeguards to make sure the data has not been corrupted from the source that sent the data, and also to make sure that any missing packets are retransmitted. Virtually the entire computer and Internet infrastructure of the world works with TCP/IP and rarely are there any undetected errors that are not automatically re-transmitted.

If ISP's did allow the errors you claim with "cheap" gear, then no one could trust any financial transaction or account balance, the the ISP would be liable for astronomical liability for bad data that went undetected.

Most people don't realize how many intermediate routers/switches a TCP/IP packet goes through to get from the source to your home. For example, I did a TCP/IP trace on my home computer and it took 14 hops (going through 14 routers/switches) to get from Tidal.com to my home.

The whole idea that computer networks use cheap components, and that streaming hi-res audio needs something superior, is absurd. Computers have zero tolerance for undetected errors and computers would frequently crash if that were not the case.

Lol are you trying to flex your internets knowledge?

Yes data is sent in packets across the interwebs. And I am well aware of a hop is. You're confusing error correction with something else.

The ISP absolutely gives you the cheapest crap that works. You're arguing functional vs optimal. Any FMC works, but as in the same way any car will get you from point A to point B, every car will do it differently.

Question for you, what role does jitter and clocking play in those 14 hops that it took for tidal to reach your pc? Does the tracert command tell you anything about that?

All I can say is if its absurd to you then dont buy it.

 

[Mark200]

Lol are you trying to flex your internets knowledge?

Yes data is sent in packets across the interwebs. And I am well aware of a hop is. You're confusing error correction with something else.

The ISP absolutely gives you the cheapest crap that works. You're arguing functional vs optimal. Any FMC works, but as in the same way any car will get you from point A to point B, every car will do it differently.

Question for you, what role does jitter and clocking play in those 14 hops that it took for tidal to reach your pc? Does the tracert command tell you anything about that?

All I can say is if its absurd to you then dont buy it.

With all due respect, you don't understand digital communications, and TCP/IP over Ethernet in particular. The 0's and 1's cannot get changed inside a TCP/IP packet from the original source to final destination (which could be 14 hops later). If there is any jitter in the digital signal, it was in the original source (probably introduced when the signal was converted from analog to digital, or when undergoing audio "mixing" and editing), and jitter cannot be added or substracted inside the packets of a TCP/IP network. After the TCP/IP packet arrives and is decoded, and then goes through the DAC, then jitter can be added that way also.

Think of it this way. If you burn an audio CD, and place it in a secure CD jewel box, and put the CD box in a car and drive it 1000 miles, there would be no change to the digital signal on the CD while being transported by the car, and it would be the same digital signal when it arrived at its destination 1000 miles later. The CD may have jitter, but not from the car ride, not matter how "jittery" the highway was.

 

[Mark200]

I spent 38 years in software development, on large computer systems like worldwide airline reservation systems and credit card authorization systems. These systems I worked on processed 10's of thousands of transactions per second, and it is critical that the data be 100% accurate. All digital data is represented by 0's and 1's in the form of positive and negative voltage (or something similar like 1 volt for a "0" and 5 volts for a "1" or whatever else they want to use depending on the type of storage or transport method being used). I don't know where you get the idea that programmers don't know that 0's and 1's are the same as binary code.

The reason why programmers know that 0's and 1's represent binary bits, is because the bits are grouped together so that 8 bits form a byte, which can have an octal value or a hexadecimal value, depending on whether one is working on a octal computer or hex computer. Most computers these days are hex, but some scientific computers are octal. Each byte has an equivalent ASCII code that the programmer works with.

Take a look at this ASCII chart to understand what I am talking about:

https://web.alfredstate.edu/faculty/weimandn/miscellaneous/ascii/ascii_index.html

Having owned one of the first consumer CD players on the market from the early 1980's, I know exactly what jitter is, and what it sounds like.