Switches and cables have no effect on the data stream as long as they are not faulty, trust me.
Not sure why people claim they hear things differently? I think it is just a waste of time.
Absolutely. But electrical noise (RF) superimposed on the digital stream then getting through the rendering process and into the DAC can cause changes to the analogue audio, through some modulation effect. Some DACS are more susceptible than others, and conceivably the effect on the sound may depend on the DAC design, so experience with one may not be the same with another. And of course it depends on the rest of the network.
Different switches and cables may have different noise pickup, or noise filtering, and so conceivably could change the audible effect. Perfect blocking of the noise, including perfect isolation of connection to ground to prevent ground plane modulation (apparently a significant factor that some DAC designs go to great lengths to minimise) would give the reference signal. Anything else is a deviation of some form - and of course some people may prefer the modified sound, just as some people may prefer the sound of notably âcolouredâ speakers or untreated listening rooms.
You are going round the circle again, @Innocent_Bystander.
A streamer receives an audio data stream via an ethernet cable or wifi, buffers the data to a certain size, processes, upsamples the data and then converts them to analogue signal. Where do all the electrical noises come from? The data are just bits and bytes.
Bits are not digital (only either on or off) if you look at them closely enough.
They are analog waves of electricity.
And they exist in a messy bath of such waves.
Thatâs where the noise is.
(Not ânoisesâ, as you put it).
Noise is all the other movements of the local electromagnetic field Apart from the signal.
I would have thought you might have learned this at MIT!
Rf noise picked up at some point or points in the network, carried together with the pulses representing the data. The pulses get through quite happily transferring the file info to be assembled into a real time digital stream by the renderer (bits indeed are bits): but the low level AC signal can also get into and through the renderer, to cause modulation effects in the DAC process. Various attempts by manufacturers are made to block it, but received wisdom suggests that extremely low levels apparently can still cause an audible effect. Tge exact nature of the effect may depend on tge specific conversion process, and how audible it may be is likely to depend on the rest of the system including the room, and how attuned the listener is to that type of effect.
The only occasion I have heard anything associated was nothing to do with a network (I donât stream music across a network), but a rendered signal from a non-specialist computer USB output, into Hugo1 DAC: To me it sounded veiled, which effect was removed by passing the signal through an isolator. Hugo1 is known for being particularly susceptible, not having inbuilt galvanic isolation, and the RF levels likely were high given the source. Changing DAC to Dave and there was no longer a problem - but I understand Dave has been designed with particular focus on effective blocking of RF. I gather also that ground plane modulation is an issue that affects possibly all DACs, and can only be completely resolved by complete isolation from ground, including that caused by a mains supply (i.e. using battery power).
As I stated, this is received wisdom - I have not made a thorough study of the subject, but this, or something along these lines* makes sense to me, because indeed the claimed effect of these switches and cables canât be affecting the actual digitised music or the whole use of networking for computing wouldnât work. It fits with my own experience, and also explains why so many people claim to hear differences with network components even when care is taken to remove bias. The latter point is important of course for anyone considering if the reported effect might be relevent to their system, and equally important is identifying the specific streamer or renderer/DAC used, a description of the network connected, and even the rest of the system, as without context it is simply a rather random anecdote.
(*IIRC it has been suggested that the noise may affect clock stability, but I may be mixing two things there.)
An analogy may help.
Words are interpreted at the level of meaning. They are semiotic/semantic entities.
The same word can be written in sand with a stick, printed in a book in ink on paper, appear on a screen in pixels, and exist as a soundwave from mouth to ear.
These different forms of the word are the same message carried by different media.
They are analagous to the bits that code the music and the medium in which those bits are instantiated (electrical waves in the electromagnetic field).
If I write the word âMcLuhanâ on a piece of paper with a fountain pen someone who knows English should be able to read it.
If the ink gets smeared a tiny bit it can still be read, and is still a word that transfers a particular meaning.
If it gets smeared more and more, it will get harder to read.
Eventually it will no longer be a word and will not transfer any meaning at all. It will just be a mess of ink.
The analogy shows us that there is a difference between the data that comprise the music file and the medium in which those data are encoded (e.g. electrical waves or optical pulses).
The DAC will get the bit perfect message either from optical or electrical media.
But as the DAC converts that message into analog waves of electromagnetic energy to send it on into the hifi the timing and magnitude of the waves can be distorted again if thereâs a lot of other EM noise (e.g. RFI) flooding through the system.
Are you saying that switches and cables have no effect on sound quality? Because if you are I find your closed minded ignorance difficult to comprehend but more of a problem is you are stating it as a fact. I have proof to say it does make a difference and it is good if people decide for themselves whether these changes warrant the expenditure.
Here we go!
Been using a EE switch for just over a week and in my system a worthwhile upgrade over the Netgear switch I was previously using.
It would be interesting to know whether a Cisco at a tenth of the price would have yielded a similar improvement. When I got mine, nigh on four years ago, it was audibly better than the Netgear that preceded it.
Yes, here we go again.
I am out of here.
PS: https://www.linn.co.uk/technology/katalyst - This is a simple explanation as to why I say what I said earlier. When I have more time, I can explain more clearly and detailed why all you need is a reliable, stable network.
Itâs not an explanation.
Itâs a load of marketing guff, with lifestyle pictures and DSP and multi-room and surround sound sales pitches.
Yes, it is, but I hope that you understand it from the architect point of view?
Please enlighten me.
Nope! You help yourself. , since you declare yourself to be âtheâ expert (hint: see your own post earlier).
Another hint: how do your superior cables, switches and routers undo all the noises generated over the bits & bytes transported over thousand/millions miles over the internet?
One of the problems that an audio streamers may face is that it can be bombarded by all kinds of network datagrams if the your LAN does not handle IGMP properly.
Not sure how an architect has anything to do with whether or not network components can add noise or filter out noise, or whether the DAC is susceptible to electrical noise introduced with the digital data?
But regardless, the Linn advert seems to suggest their DAC is immune to noise, presumably including all aspects of RF noise that may be introduced via the network, and presumably including the troublesome ground plane modulation. As I indicated in my answer to your question about what noise, DACs can be very variable in their susceptibility, and some through their design, which may include specific filtering and isolation, may be immune, or as close to immune as makes no audible difference to any given listener in any given system. If the Linn product is one of those, then that is great, and different cables and switches will make no audible difference, as no doubt numbers of people find with some other DAC products.
But for people who have got DACs that are susceptible then network components may have an audible effect, and the more susceptible the DAC then the more significant may be any audible effect, which to me explains the observations made by people posting and in the various switch and cable threads, or at least by those who have taken care to assess as objectively as possible. If I have read correctly it would seem that the latest generation of Naim DAC products have improved rejection compared to the older generation, which has to be good. And of course one personâs inaudible difference is anotherâs major effect, simply because their ears are attuned to whatever the difference might be, however small.
The one question always in my mind reading peopleâs claimed preferences is which products actually maximise the cleanliness of the data and hence minimise the modulation or the distortion of the analogue signal, giving the most accurate reproduction, and which simply modify in a way that the individual simply finds pleasing. One description I have read specifically about ground plane modulation is that it can cause the impression of a âbrighterâ sound, which some people think of as an improvement, but is actually a degradationâŚ
Thanks @Innocent_Bystander. I think your post is very reasonable from a laymanâs point of view. I would give you a B- for technical inaccuracies. However, it is still very much better than those posts from other gentlemen here, who always scream loudly from the top of their throats that they pay $$$,$$$$ for their expensive cables, switches, LPSs, etc so their HIFII system are vastly improved, sound superior to the systems that do not invest in those.
Thatâs not the way I read it at all. It seems to me that reconstruction of an analogue signal is not perfect, and that Linn are admitting as much in their sales pitch (as I believe would Naim if you asked them.)
For example, Linn say:
âcreating the analogue signal with greater precision and less distortionâ so they are not claiming absolute precision and zero distortion.
âfar lower distortionâ not zero distortion.
âprepares the digital signal for conversion with greater accuracy, minimising errorsâ not absolute accuracy and zero errors.
âhigh precision clock with its own independent power supply ensures greater timing accuracyâ not perfect timing accuracy.
âprepares the delicate analogue signal for transmission out of the device, making it less susceptible to noise and degradationâ so despite Linnâs best efforts, still still a little bit susceptible to noise.
In fact the article is full of comparative rather than superlative terms, never mind absolute claims of perfection in terms of mitigation of the imperfections of the signal that carries the data. Nice to see that they are being honest!
Maybe âimmuneâ is too strong a word, perhaps I should have said they appear to imply that they have minimised susceptibility to being affected by electrical noise arising from the network. I have no idea myself how true that may be, or whether it is just marketing speak: for all I know Naimâs might be able to claim similarly for their latest streaming products. I do recall at least one person on one of the cable threads saying they use a Linn product and could hear no difference between different ethernet cables, but it was a small sample.
As for the holy grail of perfection in reproducing precisely the original sound, it may indeed not be perfect but arguably digital may able to get closer than vinyl - but certainly many do not agree, and that is a subject of another current thread and not for here.
When I got my first Cisco 2960 two years ago it was audibly better than the D-Link switch that preceded it.
But Iâve had an EtherRegen in my system for a few weeks and my three (!) Cisco switches are now consigned to the spare parts cupboard. Iâd say the SQ uplift (with a NAC 272) going from Cisco(s) to ER is greater than with basic consumer switch to Cisco 2960.
Sorry, no experience with the EE.