[adrianm]

Just in case anyone is interested

 

[Somafunk]

Utter bull, marketed by grifters, bought by fools

 

[Clive101]

Just in case anyone is interested

Utter bull, marketed by grifters, bought by fools

"Perhaps", this model is rubbish I have no idea all I can say I have an AfterDark switch fully modded all I can say with Dave it makes a difference as does a decent server.
Soon I be taking the AfterDark to a well know member for comparison on their Dave system.

 

[motu81]

My personal experience is that the switch can make a huge difference, much bigger than makes sense. I expected it to be total marketing bull and not make any difference, but I was clearly wrong. My experience is that the biggest difference comes not from the switch itself but from the power supply to the switch, much like SJ ARC6 DC4 for Dave. Best PSU for switches I tried out was Ferrum Hypsos. Then probably comes the network cable and there I fully agree with Taiko audio that the best network cable is a plain UTP cable (unshielded), very cheap and clearly beat super expensive network cables. As for the switch the best one for me so far is the SOTM SHN-10G with their external clock. Many audiophile switches I tried out were about the same as ordinary ones, so definitely not all make a difference.

Also, I think it is not so simple as to presume that the switch only affects streamed content, as local files also sound better. If I was to guess it is probably connected to how the network signal affects the clock and timing of the source?

 

[adrianm]

My experience is that the biggest difference comes not from the switch itself but from the power supply to the switch,

This is the only thing that makes sense regarding networking. Re-clocking digital signals is bs since the source buffers the data.

 

[Reactcore]

My personal experience is that the switch can make a huge difference, much bigger than makes sense. I expected it to be total marketing bull and not make any difference, but I was clearly wrong. My experience is that the biggest difference comes not from the switch itself but from the power supply to the switch, much like SJ ARC6 DC4 for Dave. Best PSU for switches I tried out was Ferrum Hypsos. Then probably comes the network cable and there I fully agree with Taiko audio that the best network cable is a plain UTP cable (unshielded), very cheap and clearly beat super expensive network cables. As for the switch the best one for me so far is the SOTM SHN-10G with their external clock. Many audiophile switches I tried out were about the same as ordinary ones, so definitely not all make a difference.

Also, I think it is not so simple as to presume that the switch only affects streamed content, as local files also sound better. If I was to guess it is probably connected to how the network signal affects the clock and timing of the source?

As usual its again down to a groundloop bringing RF with a shielded cable being worst than unshielded. It makes no difference if local or online content is played then.

I have a cheapo TPlink mediaconverter at my server side. But powered from the server's own 5vdc then optical to a SFP in my switch.

I ordered a new mainPCB with onboard network SFP to cleanup my rig.

 

[motu81]

It’s not as simple as just psu and ground loops, as the sq delta between the sotm and a tplink is huge, even though both benefits from a good psu and from utp cables.

I wouldn’t underestimate the unknown unknowns. I’m fairly sure that the clocking plays a big part and that just assuming buffering solves all clocking and jitter issues is way too optimistic. For one, I think it’s probably harder than one might think to implement buffering well and even if that is done the source’s handling of a jittery signal could very well affect the stability of its own clock.

For the sotm switch the sound was much more natural using the sotm clock than using a totl esoteric g1x clock which sounded harsher.

Another tweak I noticed was changing the fiber modem from the operator to a higher grade huawei 10gb fiber modem and using a good psu. For the modem strangely enough I preferred using the sotm psu over hypsos.

 

[LucyWu]

Buffering processes are not a problem. Big engineering has solved buffering and shrunk the solution down into chips and you don't need to worry about them. They are part of data delivery in networks.

Jitter in packet data transmission systems is meaningless. Providing what you assume are "better" clocks in packet data products is a fools errand. Data networking is engineered on a mass scale to work. Changing an oscillator in a packet network will make no difference.

There are no "unknown unknowns" in data networking. The process of sending files broken down into little chunks, boucing them around the world and reassembling the little chunks into a file is entirely understood and incredibly robust.

Even power tweaks on a network product is going to be marginal at best and most likely completely irrelevant also.

There are possibilities for jitter at the output of the buffer in the dac but properly designed dacs solve this and buffer under-runs are possible on congested networks but fiddling with the components of a network switch or router is not going to solve that.

There really is no link between timing of bits arriving over a network interface and the audible performance of a dac unless the dac engineers have not implemented the network interface to an acceptable standard. As most of this is off the shelf componentry it's not really an issue. Grounding schemes and stray rf are all you're left with.

What you hear is not improved engineering but proof - whether any of us like the conclusion or not - that our brains are more susceptible to tweakery than the hardware.

 

[GuiltyRocker]

Buffering processes are not a problem. Big engineering has solved buffering and shrunk the solution down into chips and you don't need to worry about them. They are part of data delivery in networks.

Jitter in packet data transmission systems is meaningless. Providing what you assume are "better" clocks in packet data products is a fools errand. Data networking is engineered on a mass scale to work. Changing an oscillator in a packet network will make no difference.

There are no "unknown unknowns" in data networking. The process of sending files broken down into little chunks, boucing them around the world and reassembling the little chunks into a file is entirely understood and incredibly robust.

Even power tweaks on a network product is going to be marginal at best and most likely completely irrelevant also.

There are possibilities for jitter at the output of the buffer in the dac but properly designed dacs solve this and buffer under-runs are possible on congested networks but fiddling with the components of a network switch or router is not going to solve that.

There really is no link between timing of bits arriving over a network interface and the audible performance of a dac unless the dac engineers have not implemented the network interface to an acceptable standard. As most of this is off the shelf componentry it's not really an issue. Grounding schemes and stray rf are all you're left with.

What you hear is not improved engineering but proof - whether any of us like the conclusion or not - that our brains are more susceptible to tweakery than the hardware.

So basically messing around with switches and clocks is a waste of time and money.

 

[motu81]

I have tried countless tweaks, the majority of which either made no difference or made a negative difference. Among the no/bad ones were many that plenty of audiophiles were raving about. The switch is one of the last areas I tried since I literally expected nothing, due to the reasons mentioned, but the results were definitely surprising. If you have not actually tried it out, I would strongly encourage you to do so before arriving at a final conclusion.

I can only attest to it actually being one of the areas that made the biggest difference, roughly on par with ARC6 DC4 Dave vs normal Dave. Trust me, this is definitely not because I was expecting an uplift.

You can look at my gear list. I'm quite serious about this and have tried out a big portion of everything out there in terms of Headphones.

 

[griff500]

Even power tweaks on a network product is going to be marginal at best and most likely completely irrelevant also.

Marginal might be sufficient to be worthwhile for some people. It’s certainly not the first component I would recommend upgrading.

I think there is already a topic for this sort of discussion and it might be better there than the Dave thread?

 

[LucyWu]

Griff is probably right that this isn't the place for a discussion on either effectiveness of non-Dave tweakery or the efficacy of network kit tweakery. I have no skin in the game - I don't stream. Not because I don't think it can sound good, but because economically I don't believe artists get adequate financial support from this form of distribution of their product.

 

[adrianm]

Well i appreciate it was moved to another thread instead of deleted