When discussing digital audio, we can all agree that its understood the data is bit-perfect and any changes in sound is not because the data itself is altered. An easy way to test for bit-perfect data is to run DSD over DoP (it will sound horrible if bit faults).
When it comes to ethernet, its electronic noise that can affect the sound, not bit-faults (although re-sends for faulty packages will cause a little extra electronic noise). For SPDIF, there is also jitter to consider (but then, for toslink, electronic noise is not an issue).
The reason electronic noise can affect the sound even for bit-perfect data is that all DACs have an analog stage, where even very small changes in output voltage will be audible (zero to max volume is represented by a few volts after all). Its here that electronic noise can have a negative effect of the sound.
@Magnus (Not sure whether you were replying to my post(s) or the topic in general.) Assuming by âelectronic noiseâ you mean something to do with the extra work being done by the Ethernet/USB hardware and/or software, in principle I agree with you. In practice, the effect of such noise must surely vary (and might be completely negligible) depending on the design and engineering of the DAC.
That seems debatable, depending on the protocol being used. Happy to be corrected, of course.
This is the main point of misunderstanding. Commercial audio and video streaming services like Tidal, Spotify etc. do not use protocols that will retransmit lost/damaged packets.
I find that very hard to believe (not saying youâre necessarily wrong) as it would cause very obvious drop-outs, which would result in a lot of unhappy customers - not just âaudiophilesâ. If that were true, it would not be remedied by changing your router or Ethernet cables in any case.
Oh my, have I already said how much I love this thread? It provides some much needed light entertainment while Iâm working on something tedious.
I refer to @thumb5 's excellent clarifications on data transmission and from the replies I still conclude that some people believe extra noise is carried along with data.
Now I can accept a claim of EM interference influencing the analog output of a streaming endpoint. Itâs just that switches, routers and ethernet cables do not come into play here.
Switches and routers for obvious reasons: they are physically removed from the streaming endpoint. RJ45 connectors are part of the streaming endpoint. I can accept that there may be some EM interference from the current needed to power the RJ45 connector, but as there is current running through all the circuitry in an endpoint, youâd think that a halfway decent manufacturer has provided shielding and isolation.
The cable has nothing to do with this whatsoever. It just delivers bits to the RJ45 connector. Thatâs it.
Thatâs why data streams are buffered before the DAC and sophisticated interpolation algorithms trying to fill in the holes.
And 2020 the Internet has much more bandwidth than 10y ago = far less audible drop-outs.
Ethernet cables, like all digital cables, are analog cables where +V pulses is interpreted as a 1, no cable can deliver âbitsâ directly.
And I posted some measurements a while back that show how easy ethernet can pick up RFI from the outside, just by being close to for example a monitor. Ethernet is no more immune to RFI than any other electric cable, and even if shielded some RFI will slip through, and some electronic noise will be carried along from the ethernet switch.
Its possible to quite easily improve this by using media converters and fiber, but you still get noise from the last media converter (as well as any noise thats picked up from the outside). But it is an easy and cheap tweak that results in an audible improvement:
Obviously, the data is still delivered with this solution, but up-stream electronic noise is removed by the fiber.
Yes. So to the question âDo router and ethernet cables affect sound quality?â the answer will be:
Yes, it could, but only in a negative way.
The SQ cannot be âliftedâ, like Melco is claiming.
And my point in this thread is that standard priced equipment can achieve 100% acceptable transport, but if you get noise that affects your DAC or other equipment, you must of course do something.
Thatâs more of a philosophical question, but I like to see all cables, including analog signal cables, as something that remove sound quality to some extent, the less the better. But compared to bad cables, good cables can be seen as improving the sound (although in truth they are really removing less sound quality than the bad ones).
That depends on what you mean by acceptable. If all you want is data to be delivered properly, then yes pretty much anything will do. But if you also want to maximize sound quality than you need some specialized gear with linear voltage regulators etc (switched electronics is bad for sound quality).
Acceptable for me is no audible noise whatsoever - playing a silent track and turning the amplifier way over my normal highest listening volume or to the max.
All kind of really detailed hi-tech measurements exist already, there are available in the following link:
There is no measurable difference in noise, jitter, dynamic range, whatever, at the DACâs outputs, by changing the ethernet cabling and even ethernet equipment. If it works, it works.
But it is beautiful and quite enjoyable the idea of fixing an undetectable problem with an unmeasurable and unexplained device.
How big is the difference you think you perceive? How do you know itâs really there? How do you know youâve achieved âmaximized sound qualityâ? How can you be sure things have actually âimprovedâ?
Thatâs a good start, but it only means you have little interference when no music is playing, which does not automatically mean you have little interference with music on. Also, noise can affect sound but not affect silence.
Just to clarify: electronic noise might still exist and affect sound, even if it makes no âhissingâ noise from speakers. For example, electronic noise will have a negative effect on crystals (clocks), which will produce more clock-jitter, but this only happens when music is playing and even then it donât make any âhissâ noise. It will however remove some details and dull the sound.
A couple of days ago I got a long PM from someone with speakers for 50,000 dollars and cabling etc. that cost him more than five times as much. I still canât believe itâŚ