True?? An "optical break" in network cable can improve SQ

[Gorgonzola]

A member on another forum claimed he improved sound quality by insinuating an "optical break" in his network cable, (between NAS and streamer I believe).

The "break" consisted of a pair, back-to-back, of these TP-Link MC100CM electrical <=> optical converters in his home network cable. He insists that he and a buddy could hear an SQ improvement in 48 to of 50 trials using blind testing. The "optical break" was to reduce network noise to effect the improvement. Assuming he's completely honest, which I'm not doubting, what conditions could cause his positive result?

FWIW, another guy on the forum is using this $2800 Melco S100 switch claiming SQ improvement, (but no mention of blind testing).

 

[Taddpole]

"Assuming he's completely honest, which I'm not doubting"

I'm not joining you in that assumption.

 

[BDWoody]

Assuming he's completely honest, which I'm not doubting, what conditions could cause his positive result?

Testing methodology gets my bet.

I don't doubt the sincerity of most who make whatever claims they make. I simply doubt they conducted a valid test.

 

[NiagaraPete]

A member on another forum claimed he improved sound quality by insinuating an "optical break" in his network cable, (between NAS and streamer I believe).

The "break" consisted of a pair, back-to-back, of these TP-Link MC100CM electrical <=> optical converters in his home network cable. He insists that he and a buddy could hear an SQ improvement in 48 to of 50 trials using blind testing. The "optical break" was to reduce network noise to effect the improvement. Assuming he's completely honest, which I'm not doubting, what conditions could cause his positive result?
View attachment 181688

FWIW, another guy on the forum is using this $2800 Melco S100 switch claiming SQ improvement, (but no mention of blind testing).

Is it April 1st already or is the guys name P.T Barnum? Networks don't have noise.

 

[Honken]

In pathological cases I guess it could break a ground loop, but I'm no EE so I do not wish to speculate on the exact circumstances as to when that could be true. Wi-Fi would be a more sensible option for most home users to accomplish the same thing.

 

[alpha_logic]

This is total bullcrap, one big flaw with the whole 'magical network switch' theory is the fact that the file is actually buffered by the OS, which means on repeated playback it's not even streamed over the network, but played back from the host system's cache/buffers. Yes - it depends on the protocol used (cifs, nfs, dash, whatever), the device used etc - there are exceptions to this rule - but in general that's what happens, and I've seen none of these 'tests' actually account for the host cache.

 

[morpheusX]

Sorry, couldn't resist

 

[somebodyelse]

In pathological cases I guess it could break a ground loop, but I'm no EE so I do not wish to speculate on the exact circumstances as to when that could be true. Wi-Fi would be a more sensible option for most home users to accomplish the same thing.

The signals are transformer isolated on wired networks. Most of us use unshielded network cables, so the ground loop is already broken. If they were using a shielded network cable then the ground loop is a possibility.

 

[antcollinet]

The signals are transformer isolated on wired networks. Most of us use unshielded network cables, so the ground loop is already broken. If they were using a shielded network cable then the ground loop is a possibility.

If there is a ground loop, it can be connected from Nas to streamer to amp, if shielded cables are being used (audiophiles might). If this is broken between nas and streamer, then it could well eliminate ground loop noise on the analogue connection between the streamer and the amp.

Feasible that it can result in an improvement in that case.

 

[radix]

In pathological cases I guess it could break a ground loop, but I'm no EE so I do not wish to speculate on the exact circumstances as to when that could be true. Wi-Fi would be a more sensible option for most home users to accomplish the same thing.

All ethernet has galvanic isolation. Ethernet (usually) does not even have a ground line, it's just several balanced pairs. There are some special-use shielded Ethernet cables, but they need special gear (you ground the patch panel and attach the shield there, then float it at the end device) and are pretty uncommon.

I would not add a wifi bridge just to isolate, use those optical couplers. But I don't think its at all needed.

Optical in ethernet is meant for digital noise isolation, such as in manufacturing floors, for long runs, for very high data rates, or in some rare cases for security.

 

[radix]

Amir tested an isolator. If he really, really, tried he could see some super tiny improvement.

 

[Aldoszx]

Snake oil, I suppose.
The only logical explanation (not taking into consideration audio and referring only to a network connection) could be a copper wire longer than 100 meters. In that case, this "optical break" means in fact an optical network path ended with copper wires via media converters which surpass the length limit of cooper network.
I don't know what is the symptoms of using a longer than standard network cable for audio network streaming but I think I could try.

 

[Honken]

This is exactly why I don't speak in absolutes on these matters.

 

[AudiOhm]

TCP/IP is the Ethernet protocol. The music at this point is just another file, how does it know how to improve just the music files?

The device you have listed is for long distance transmission. Network switches have built in isolation...

Some videos of interest.
https://www.youtube.com/user/mnoble247

https://www.youtube.com/channel/UCFxyjxLmyDdW8-oC9qjV38w

Share the link.

Ohms

 

[JeffS7444]

There's no need to speculate: If they can share sound files with and without the converters in place, we can check them out for ourselves.

 

[Berwhale]

Is it April 1st already or is the guys name P.T Barnum? Networks don't have noise.

Of course they do at the physical layer. But there are multiple layers of abstraction between the noise affecting the electrical signal and the bytes that get reconstituted into the audio data on the receiving equipment. The Open Systems Interconnection (OSI) model of networking has 7 layers of abstraction: Application, Presentation, Session, Transport, Network, Data Link & Physical*.

Here's how the layers of abstraction work when two applications (which would be the NAS and the Streamer in this case above) talk over a network...

This is how the logical layers in the OSI model (roughly) relate to actual networking standards that you may have heard of...

*I learned the layers of the OSI model in early 90's with an aide memoire of my own invention:

"All Priests Should Take No Lustful Pleasures"

...and I have never forgotten them

 

[Berwhale]

TCP/IP is the Ethernet protocol.

Actually, no. TCP is the transport layer (Transport Control Protocol), Ethernet covers the physical and data link layers.

 

[mansr]

There are some special-use shielded Ethernet cables, but they need special gear (you ground the patch panel and attach the shield there, then float it at the end device) and are pretty uncommon.

Most Ethernet jacks have shields that will connect to shielded cables, no special equipment needed.

 

[AudiOhm]

Actually, no. TCP is the transport layer (Transport Control Protocol), Ethernet covers the physical and data link layers.

You are correct, incorrect wording by me...

Ohms

 

[radix]

Most Ethernet jacks have shields that will connect to shielded cables, no special equipment needed.

Well, if you want to avoid ground loops, you need to do something.