Does connecting a streamer to Ethernet make much difference compared to wifi?

[Peter Keen]

When I had an RME DAC it was easy to test the reliability of streaming files from various sources on my home network via WiFI, ethernet direct from the router and ethernet via power-line adapters All routes were able to transmit bit-perfect test files up to 192/24. They also sounded the same. I am well aware that the advocates of network tweakery will cite “noise” of various forms being the demon to be tamed in wired streaming, not the accurate delivery of data but to my ears in my systems I could not distinguish between wireless and wired transmission. I say go with what is most practical/stable.

 

[Jubal]

It isolates the streamer from the router.

Why is that necessary?

 

[tuga]

Why is that necessary?

To reduce the amount of RF noise travelling along the network which affects clock and D/A chip performance.

 

[vacuum_tubes]

To reduce the amount of RF noise travelling along the network which affects clock and D/A chip performance.

absolute gibberish

 

[tuga]

absolute gibberish

And you are?

 

[vacuum_tubes]

And you are?

Someone who understands the difference between foo and reality.

 

[SteveG]

I tend to wire my streamers where possible (sometimes using powerline adapters) as I've found it more stable - especially in my cottage which has thick walls and wi-fi struggles a bit.

I've not noticed any sound quality difference!

 

[narabdela]

There are other threads that talk about this. Very contentious!

'Contentious' is putting it mildly. They blow cable threads out of the water.

( Sorry, couldn't resist the meme).

 

[Fergus]

'Contentious' is putting it mildly. They blow cable threads out of the water.

( Sorry, couldn't resist the meme).

Yes, I thought it would set this thread up for inflammatory comments and ruin it for the OP.

 

[adamdea]

My music room is in a different part of the house hence wondering if it might be worth trying to run a cable from my router. Internet is fibre and I use BT Discs to relay the signal currently.

Thanks

Jez

If we are being honest with ourselves, what does this question actually mean?

-Worth trying in the hope of what? If you have a problem, it is possible that it might solve it. But do you? people often find wired connections more reliable, especially if you have a separate music server and renderer(s)-I have my LMS server pi and my NAS wired, but my endpoints are wireless. I followed this advice years ago and it works. It also makes sense because

• the server needs two data hops (in from nas/internet and out to renderer)
• the server and nas can be tucked away anywhere but the renderers have to be where you are listening.

but if it already works fine then it works fine.

Does the question in fact mean "if I try doing this is it possible that it will solve a problem I didn't know I had?"

If the question is -is there a hope that after doing this I might think that it sounds better then the answer is "yes". But then the answer to that question is always "yes".

 

[adamdea]

I think there might be a market for deliberately flawed products which can have their flaws solved by buying upgrades. You could sell the device with eg a PSU designed to inject proper audible amoiunts of intereference. Then you could buy a replacement for say £500 which doesn't.

This would be great because then hifi component upgrade paths would work the way people want them to. We could start with a router which actually does inject noise via ethernet and a dac designed not to reject it. Perhaps the easiest thing would be to have analogue components labelled as digital. Maybve take spotify put it through a DAC then route the dac through a compressor to the amp. The dac and compressor could be in a box marked "level 1". Then you could pay 5k for a box marked "level 2" which just wouldn't have anything in at all.

Any investors out there?

 

[JezmondTutu]

If we are being honest with ourselves, what does this question actually mean?

-Worth trying in the hope of what? If you have a problem, it is possible that it might solve it. But do you? people often find wired connections more reliable, especially if you have a separate music server and renderer(s)-I have my LMS server pi and my NAS wired, but my endpoints are wireless. I followed this advice years ago and it works. It also makes sense because

• the server needs two data hops (in from nas/internet and out to renderer)
• the server and nas can be tucked away anywhere but the renderers have to be where you are listening.

but if it already works fine then it works fine.

Does the question in fact mean "if I try doing this is it possible that it will solve a problem I didn't know I had?"

If the question is -is there a hope that after doing this I might think that it sounds better then the answer is "yes". But then the answer to that question is always "yes".

I just wondered if there would be an improvement in SQ if there is more interference on wifi that’s all.

 

[Phyztech]

Btw, there’s no conflict in running both wifi and ethernet at the same time - just put 2 routers in bridge mode and run ethernet from the one that can be placed closer to your system.

 

[Phyztech]

We could start with a router which actually does inject noise via ethernet

They all do that as any device with a power supply

 

[Amber Audio]

I just wondered if there would be an improvement in SQ if there is more interference on wifi that’s all.

No improvemnet in SQ IME. Using Mesh Wi-Fi dishes like your BT or Unifi or TP Link etc I wouldn't worry about it at all. If your connection is stable and relaible - no drop outs, just stick as you are I reckon.

I'd avoid using the Wi-Fi extenders that run over the mains - Powerline - they can inject noise/clicks etc. but not the Dish based systems.

 

[adamdea]

I just wondered if there would be an improvement in SQ if there is more interference on wifi that’s all.

let's be honest- you want to buy something don't you?

 

[vacuum_tubes]

I just wondered if there would be an improvement in SQ if there is more interference on wifi that’s all.

No improvement in SQ but wired is generally more stable/reliable than WiFi. That's not to say it's possible to make wifi just as reliable, it's just more complicated or harder for people to achieve at home.

The only time I'd strongly recommend wired over WiFi is with Roon as this flies in the face of convention.

Most streamers use decent buffers so network performance is not a critical issue, i.e if packets are lost they just get retransmitted and fill the buffer. Most streamers will store an entire song in the buffer. My old streamer used to store around 3-4 songs so you could disconnect the ethernet and it continued playing for 10-15 minutes.

Roon uses tiny buffers in order to keep the multi-zone playback in sync, the downside is that any poor network performance is manifested almost real time with glitches and stutters (and often with the stream stopping completely) It boggles why Roon don't use proper buffers when playing to a single endpoint and then small buffers when using multi zone. Then again the Roon developers are weird and seem to design everything in the worst possible way on purpose.

 

[adamdea]

Btw, there’s no conflict in running both wifi and ethernet at the same time - just put 2 routers in bridge mode and run ethernet from the one that can be placed closer to your system.

Yes that way the thing that injects noise into the system can be closer to your streamer. Plus you have got to buy something.

They all do that as any device with a power supply

Well yes, but a really appreciable (potentially audible) amount. You know, "noise", the sort of thing which makes a competently designed dac unable to reconstruct the data. I've been reading about this stuff in hifi mags for years. The sort of thing which might be removed by a PS Audio noise harvester.

 

[Hcanning]

I think there might be a market for deliberately flawed products which can have their flaws solved by buying upgrades. You could sell the device with eg a PSU designed to inject proper audible amoiunts of intereference. Then you could buy a replacement for say £500 which doesn't.

It wouldn't surprise me if there was some element of truth in this. Surely for a modern streaming device to have the output of its audio noticeably affected by a tiny bit of spurious noise on the mains supply, it has to be a fundamentally terrible design? There's an awful lot of error-correction that goes on when a song gets transferred from a NAS to a streaming box before it gets converted to PCM and chucked out the digital outputs. These network devices deal in data rates inconceivably higher than the data rate of an SPDIF digital audio stream. You'd need a hilarious amount of interference before this was noticed on the digital output by way of pops/clicks/static/dropouts. The notion of this noise somehow acting like a tone control and making the music just sound "different" doesn't sit right with me at all.

Spend your money on amplifiers and speakers and room treatment. That's where the real improvements lie.

 

[Phyztech]

Yes that way the thing that injects noise into the system can be closer to your streamer. Plus you have got to buy something.

You can select 2nd router that is way less noisy than your main one + when you put it in the bridge mode it becomes half-router/half-switch as it only receives wifi + you can upgrade power supply + you can use switch or ethernet filter after it