In the several lengthy switch and cable threads last year it was clarified several times by a number of people that whilst the digital data being altered is simply not a viable explanation, electrical (RF) noise picked up by the network certainly can affect the resultant sound if it gets into the conversion stage of the DAC, e.g. by modulation if the analog signal, and is the most probable root cause of people hearing differences with different network components. Different network setups, the electrical environment it is in, what is connected, and including cables and switches, even switch power supplies, can affect the amount of noise picked up, or generated, or the amount of noise filtered out, and so can affect the degree of modulation of the analog signal, and hence the sound. And, significantly, different DACs will differ in their susceptibility , depending both on the DAC process and the design of the overall unit (e.g. inbuilt filtering), likewise the renderer/streaming stage before the DAC, whether in the same box or separate.
So it is to be expected that susceptibility of a streamer or DAC to electrical noise introduced by an attached network will differ according to the specific streamer or DAC product used - and entirely conceivable that a manufacturer could design theirs to effectively block and be immune to introduction of noise via the network connection and so be unaffected by whatever network cables or switches are used.
That is the recommendation along side wifi of course these daysā¦ the world has moved onā¦ but admittedly there are more variables with wifiā¦ so from a support perspective Ethernet is easier to assure hence recommended.
I think however reading these forum pages there is a confusion and a ācrossing of the beamsā in this area of home networks āoptimisationā. Many of the issues discussed on these pages, are nothing to do with the ādata networkā but rather conducted electrical noise coupling into audio systems and the optimisations are more about noise shaping. Remember products work as a system as a chainā¦ in the real world each product canāt have a perfect boundaryā¦ and understanding this is good engineering practice.
On the actual network sideā¦ there is scope for optimisation on combined streaming/DAC equipment. If you can hear the difference between WAV and FLAC on your streamer equipmentā¦ then almost certainly you will benefit from network optimisation such as IGMP snooping increasingly for home networking, due to multicast group addresses being increasingly used in our homes. Say 10 years ago, Iād say this was mostly irrelevant. Increasingly consumer network equipment includes this as standard apart from the very cheap and basic devices.
Certain network switches can also help methods used by streaming applications and their discovery methods, by using IGMP queriersā¦ higher end consumer routers like the BT SmartHub2 have these built inā¦ so all plug and play.
With the older Naim streamers there was also the variability of media streamer server types (UPnP servers), based on their inter frame timing consistency .I shared my own findings here with Naim when they were developing the new streamersā¦ as far as I can hear (and see) the new Naim streamers are agnostic to this nowā¦
So there is always things underneath the coversā¦
However the main thing I have learnt is to decouple the streamer transport wherever possible from the DACā¦ and things become more consistent from a sound performance perspective.
Although decoupling is part of Naimās DNA for much of their portfolio, itās not yet for their streamersā¦ hence why we see pages and pages of network tweaked etcā¦ I suspect this might change at some point along with the adoption SFPs to allow customer choice of fibre or twisted pair.
I discussed the same point with dcsā¦ switches and lan cables, and their recommendation was a basic, unmamanged cisco gig switch and good quality lan cable i.e. not lan cable with uses 26swg (or worse) wire and/or alluminium.
It is no surprise really that dcs say this - the guy from PS Audio has a YT vid saying the same. Iām happy with my 8switch and I know it makes a difference in my rig, but it may not for everyone. Is it placebo? I hope not, I know all about placebo effects as I am a research psychologist. Do I know about IT and networks - yes a bit, though obviously not as much as someone who works in IT but I have been using computers since the first IBM PC came out in the 80s. The good thing is that most suppliers will accept returns if you are not happy. Speaking to my favoured UK supplier they havenāt had anyone return an 8switch. And donāt uptone audio, maker of the EtherRegen, offer a free return policy as well? So youāve nothing to lose by trying it out. These debates about 1s and 0s not being impacted by cables, switches etc will go on forever I suspect, without any resolution. Whatever makes you happy. I personally find it hard to believe folks will shell out so much money for Isotek isolation products but there you go, we all think some products are a waste of money. At the end of the day, if someone says it makes a difference to their system, you canāt really challenge that as you donāt have their ears !
If the suggestion is made that it is really about EM/RF interference rather than ābitsā, and that the streamers/DACs are so sensitive that their analogue conversion stage is significantly distorted by EM/RF coming in through the ethernet cable, then surely it must be sensitive enough to pick up anything that is emitting EM/RF in the area; cellphones, neighbourās equipment, microwaves, lighting, or even cosmic radiation. Where does it end?
We should trust that if we buy a high end streamer/DAC, that the manufacturer has shielded it enough from this type of noise to not measurably impact the output signal, at least not above the human hearing threshold. If it does then perhaps the equipment is really just poorly designed.
When designing a dac there are many many decisions to be made - and each of them will effect SQ and will also effect how EM noise changes SQ. But you canāt maximise both criteria.
Just to clarify my original post
This wasnāt meant to be another thread on switches and what it may or may not do, as plenty of them on this forum all ready.
It was more the stance that dCS have said on the matter, the fact that they also know what they are talking about and what this says about all this really.
As I found it quite refreshing that a manufacture answered the question so bluntly, than the normal no answer, or skirt around the question.
If cables made a difference to the 1ās and 0ās (or, if weāre being precise, the electrical pulses the represent the 1ās and 0ās) then if we attach a computer to one end of the system and a computer at the other, the second computer would receive different information depending on the cable configuration. That canāt be the case (else your bank account would alter if you changed the cables, to be extreme). So, the second computer can reproduce the first oneās digital output exactly.
Now, this does not mean that the signal was the same - just that it works within the protocol and the error checking can accommodate any errors. So there is one possible source of a difference - harder working error correction could possibly introduce thermal noise into the circuitry and so affect the amplification down the line. But any changes here will be minute, and I would be highly surprised if they were not negligible.
Secondly, the other source of changes in the signal could be noise from the switch itself being picked up in the amplification circuits. Since these have power supplies and so on, this could be more significant, and may make an audible difference, reduced or negated by moving the switch far away.
And finally, the changes could be confirmation bias (real enough to the listener) but not scientifically distinguishable from the alternative.
Yes, but by quoting dCSās stance on switches and cables you will inevitably invite other opinions, and it would be wrong to stifle that debate, even if we have heard much of it before.
Or, more widely:
Secondly, the other source of changes in the signal could be electromagnetic noise from the mains via all connected cables and equipment and other sources of EM noise combined being picked up and amplified and distorted in the electrical circuits of the hifi.
It ends with the people of Vortex Hifi who claim that having a mobile phone in the same room as your hifi destroys the sound unless you buy their protective technologies. Which incidentally also prevent your thrombocytes from sticking together and improves your health. I kid you not. I recommend the Vortex Hifi website for a jolly afternoon. Bring popcorn and refreshments
Edit: @litemotiv I forgot if you read German. If you read the Vortex website through a machine translation, be assured that the gibberish is most likely there in the German version as well
Its interesting that a lot of people seem to doubt dCSā motives in saying Ethernet cables make no difference.
But if dCS actually said network cables make a difference to sound quality, wouldnāt they be able to commercially benefit from such a claim by making and selling more cables?
And its not just dCS making the no difference claim, other major manufacturers also echo the same view.
Of note is that there currently is no manufacturer, and there are no scientific papers or measurements available that illustrate that EM/RF in the ethernet stack is causing audible sonic alternations in the analog output stage of a streamer/DAC.
All the evidence we have is hypothetical or anecdotal, from users on forums who report that to them there is an audible difference. It is also more prevalent on certain forums (e.g. Naim) than on other hifi forums, where i think you generally see a more sceptical attitude towards these claims.
Two other threads on the Roon forums about this topic that are worthwhile checking out:
Not sure which video you refer to with Paul. But he also say in a video that they did a blind switch of USB digital cable at a show and he was blown away by the difference.
I rarely trust sales people. They have an agenda and the agenda is rarely to tell the truth. Money talks.
