The middle switch

[Weerstandje]

In my home network are two Sotm ethernet switches. The first one is connected via the router to the outside world. From that first, standard Sotm switch (to which no hifi is directly connected) ethernet cables and a glasfiber cable fan out to other switches in other rooms. One of those is my other Sotm switch (connected with the glasfiber) that feeds my main hifi set. That last one is the better Sotm switch with a clock upgrade.

 

Now the thing I am wondering about is whether I should change the switch with the clock upgrade to the middle of my network? In my thinking my other switches (two more simple but still hifi grade ones) and the second Sotm without the better clock would benefit from the more steady stream from the first?

Or should the best switch be the last in the chain to feed my best hifi set? That’s how it is connected now, ‘cause this was the first hifi grade switch I bought. The other ones came later.

 

I know, it is a matter of trying, but it is a hell of a job getting behind everything, unplug it and plugging it in again. So maybe the theory first?

If there is theory behind it that is…

[TheFlash]

If you understand how ethernet works, you'll know a more accurate clock won't affect sound quality. In the streamer/DAC world sure, but not in moving packets of stuff around which then get unpacked by the streamer. IMHO, you'd be wasting your money with an ethernet clock upgrade.

 

Why do you have your two switches dasiychained? If you can plug both switches into your router, use one to feed audio and the other to feed non-audio. This should reduce the chances of noise from non-audio devices "polluting" the audio playback chain. Certainly it will cost nothing to try (apart, apparently, from the plugging and unplugging challenges!).

 

Report back 🙂.

 

[audiobomber]

On 10/17/2023 at 10:35 AM, Weerstandje said:

In my home network are two Sotm ethernet switches. The first one is connected via the router to the outside world. From that first, standard Sotm switch (to which no hifi is directly connected) ethernet cables and a glasfiber cable fan out to other switches in other rooms. One of those is my other Sotm switch (connected with the glasfiber) that feeds my main hifi set. That last one is the better Sotm switch with a clock upgrade.

 

Now the thing I am wondering about is whether I should change the switch with the clock upgrade to the middle of my network? In my thinking my other switches (two more simple but still hifi grade ones) and the second Sotm without the better clock would benefit from the more steady stream from the first?

Or should the best switch be the last in the chain to feed my best hifi set? That’s how it is connected now, ‘cause this was the first hifi grade switch I bought. The other ones came later.

 

I know, it is a matter of trying, but it is a hell of a job getting behind everything, unplug it and plugging it in again. So maybe the theory first?

If there is theory behind it that is…

From my experience and research, your current configuration is correct; fiber optic in the audio chain, and the best clock closest to the DAC. Nice system. 

[Weerstandje]

On 12/8/2023 at 10:16 PM, TheFlash said:

If you understand how ethernet works, you'll know a more accurate clock won't affect sound quality. In the streamer/DAC world sure, but not in moving packets of stuff around which then get unpacked by the streamer. IMHO, you'd be wasting your money with an ethernet clock upgrade.

 

Why do you have your two switches dasiychained? If you can plug both switches into your router, use one to feed audio and the other to feed non-audio. This should reduce the chances of noise from non-audio devices "polluting" the audio playback chain. Certainly it will cost nothing to try (apart, apparently, from the plugging and unplugging challenges!).

 

Report back 🙂.

 

 

Sorry for the late response 

 

Recently I talked to a fellow audiophile about the fibre-optic connection I made between my switches with sfp modules. We also discussed timing. He to said something about moving packets around, sometimes even in the wrong order and them still arriving safely at the streamer. He claimed that timing faults in the audio stream would lead to dropouts sooner then jitter effects.

I wondered if audio-streaming is called streaming because, as opposed to file transfer, a pcm stream does not come in packages. For the same reason CD-s have to be ripped to an aif- or wav-file on disk because cd’s only contain the stream and have no checksum?

 

If the later is true, streaming over glasfiber would be more prone to jitter, due to the bending of the cable giving light-reflections inside the cable, than streaming over an ethernet cable. Just like spdif is preferred over toslink as a connection between streamer and dac for instance? 

 

That made me wonder in what format music is streamed over the internet before it reaches the streamer and the dac?

 

 

[krass]

25 minutes ago, Weerstandje said:

 

Sorry for the late response 

 

Recently I talked to a fellow audiophile about the fibre-optic connection I made between my switches with sfp modules. We also discussed timing. He to said something about moving packets around, sometimes even in the wrong order and them still arriving safely at the streamer. He claimed that timing faults in the audio stream would lead to dropouts sooner then jitter effects.

I wondered if audio-streaming is called streaming because, as opposed to file transfer, a pcm stream does not come in packages. For the same reason CD-s have to be ripped to an aif- or wav-file on disk because cd’s only contain the stream and have no checksum?

 

If the later is true, streaming over glasfiber would be more prone to jitter, due to the bending of the cable giving light-reflections inside the cable, than streaming over an ethernet cable. Just like spdif is preferred over toslink as a connection between streamer and dac for instance? 

 

That made me wonder in what format music is streamed over the internet before it reaches the streamer and the dac?

 

 

 

an “internet stream” is still sent as packets of information, and the packets will still have an associated packet-protocol with headers, footers, counters, checksums etc. They arrive in your house at the modem just like any other packets of internet data. Your modem/ network/ routers will send the correct packets to your music system that requested them.

 

Your player still has to receive the packets, check them, unwrap them, buffer them etc. 

[TheFlash]

On 1/10/2024 at 3:28 PM, Weerstandje said:

 

Sorry for the late response 

 

Recently I talked to a fellow audiophile about the fibre-optic connection I made between my switches with sfp modules. We also discussed timing. He to said something about moving packets around, sometimes even in the wrong order and them still arriving safely at the streamer. He claimed that timing faults in the audio stream would lead to dropouts sooner then jitter effects.

I wondered if audio-streaming is called streaming because, as opposed to file transfer, a pcm stream does not come in packages. For the same reason CD-s have to be ripped to an aif- or wav-file on disk because cd’s only contain the stream and have no checksum?

 

If the later is true, streaming over glasfiber would be more prone to jitter, due to the bending of the cable giving light-reflections inside the cable, than streaming over an ethernet cable. Just like spdif is preferred over toslink as a connection between streamer and dac for instance? 

 

That made me wonder in what format music is streamed over the internet before it reaches the streamer and the dac?

 

 

Basically, what @krass says. Dropout on a domestic network would suggest you have a terrible network; this is not a real world problem. Your friend is right if he is telling you not to worry about your fiber/fibre connection. Unless you are doing something silly with your fibre/fiber cable, the bend is irrelevant and has no bearing on reflections. You may be overthinking this.

 

If there is any "jitter"  (timing problem) in the ethernet space, it refers to packet delay which is extremely extremely unlikely (basically impossible) in normal domestic networks, and it won't (simply can't) affect sound quality anyway; ethernet jitter is a completely different thing from bitstream jitter which can affect sound quality. Ethernet jitter is therefore a thing, but not a thing you need to worry about umless you run an overloaded corporate network.

 

SPDIF is usually preferred to TOSLINK because the latter is bandwidth limited and the optical-electrical converters are often very noisy. This has no relationship to fibre/fiber optic cable in ethernet and has nothing to do with cable bend.