Streaming over WiFi versus Ethernet Connection

[Guest rmpfyf]

1 minute ago, aussievintage said:

Anyway. The poster to whom I replied, and I, were pursuing a line of discussion that said upstream didn't matter, only noise from a cable, therefore no speed/latency/packet issue will make any difference for a properly working Wifi, and using Wifi will remove the cable noise.

 

I read his posts as simply suggesting upstream noise affects a downstream result, which is not generally untrue. 

 

3 minutes ago, aussievintage said:

Noise from cpu, don't use a computer.

 

Ha, This is an interesting one. Unless there's a leash on how it's consumed (and even then) a standard CPU in a usual motherboard is pretty compromised for audiophile work. There's better and worse out there but it's no peach.

 

So yeah, noise from CPU, don't use a computer applies  

[Dankeshon]

So use Wifi, no noise from the ethernet cable.

I don't have very expensive gear, only total 4700$ with several hifi components.

My findings is that WiFi is bad quality on most products.

The exception is cell phones and tablets...

It's not always connecting correctly, it's un- reliable. The sound is not the best, and if a crappy Ethernet cable is noisy, WiFi in a player/Dac radiate noise like nothing else.

I recommend cables.

[Dankeshon]

Nas connected WiFi to Router then connected with a shielded cable too Dac is more practical...

[Dankeshon]

Noise from cpu, don't use a computer. [emoji4]

 

Anyway. The poster to whom I replied, and I, were pursuing a line of discussion that said upstream didn't matter, only noise from a cable, therefore no speed/latency/packet issue will make any difference for a properly working Wifi, and using Wifi will remove the cable noise.

Your just adding the noise of WiFi to inside the Dac/Streamer.

 

Mostly the same EMI noise we warned about from the router is now.... Tada! Inside your streamer/Dac.

 

Nas WiFi to Router Ethernet cable to Dac is a better option.

 

[Dankeshon]

The objectist Don't believe in science....

They dont belive in jitter

They don't believe in EMI.


They joke about servers and in wall cables.....

Both in Ethernet and power.
It's mostly about EMI......

But the other part doesn't believe it exists, even when I put my meter to the router and WiFi it screams like a cat....

That's science....

[Guest rmpfyf]

Well here we go, Uptone is about to release a USD$640 audiophile switch and make my d**king around irrelevant.

 

https://audiophilestyle.com/forums/topic/38968-etherregen-we-are-getting-much-closer/#comments - starts in 2018, the last pages have beta units shipped and a price announced. One 'special' output RJ45 port, 4x RJ45 inputs and 1x SFP in. Save your biscuits if interested, or wait a year for upgradeitis to see the first ones for sale s/h on SNA.

 

Some points I can make out for the TL;DR crowd (I'm sure others will chime in):

• The output port is isolated as best possible from the rest, powered by LT3045-based circuits, etc
• Less jitter at lower speeds, hence the 'special' port is 100Mbps - the rest are Gigabit (suspect lower power requirements played a part here too) - 
• Separate PHY for the 'special' port follows Swenson's experience of best sounding switch/routers having a separate PHY per port
• Yes, it's got a special clock optimised for low phase noise used to reclock the 'special port'... and it might take a 10MHz external clock reference
• Prolly USD~$200-250 in parts at their volume, w/much component spend on getting beautiful isolation/waveform/etc out of that last port
• Will not need a LPS for best results - they use their ground-shunted SMPS to feed this thing (intended to eliminate high-source-impedance AC leakage)
• Sufficiently isolated as not to sound any better when fed optical over SFP
• No word on max MTU

 

He does mention that some SFP+ modules have very low additive phase noise.

I'd love to see someone measure OS jitter on the receiving end, it's where I think you'd see a difference.

 

Some posts covering his noise source arguments leading to the product are here.

https://audiophilestyle.com/forums/topic/37034-smps-and-grounding/page/2/?tab=comments#comment-723187

https://audiophilestyle.com/forums/topic/37034-smps-and-grounding/page/9/?tab=comments#comment-735311

 

I have no doubt Swenson and co make fantastic product, I also have no doubt I'm too cheap to spend $1k landed on a switch without a half-a**ed fight. 

 

I'd reckon most of this concerns noise leakage which is in part no issue with optical, and in part clocking and surrounding electronics. 

 

I have a spare switch with SFP identical to my main one, the oscillators inside are 25MHz and I have some spare low-phase-noise XOs and OCXOs from old projects (with power supplies)... can do the shunted SMPS... might cook up a frankenstein not-quite-as-good-but-directionally-correct-outta-spare-parts take on it in time  (This sounds more and more like 'bite the bullet and start saving, if you value your time')

 

Interesting reading though.

[wolster]

@Dankeshon

Voice your opinions by all means but please follow our guidelines and respect the views of others.

Thanks.