Yes, datawise. Soundwise? How would that happen then?
Either the packets of ones and zeroes arrive or there are dropped packets. The content of the packets is not changed by the so-called noise.
It’s not as if the EM interference mythically known as “noise” acts as if it were a computer virus. It doesn’t magically decode the streamed files, adds distortion to them and then recodes the streamed files with changed waveforms.
Besides, most - if not all - DACs have a sample buffer. If not, reclocking in the DAC would be impossible and there would be no possibility for error correction (which has been around since the first CD player).
So I fail to see why the transfer protocol would influence sound quality, with the exception of dropped packets. And as @danny stated, UDP does use checksums. There’s just no authentification protocol.
If we head into details this can get complicated pretty soon, when splitting up discussion in layers like control protocol or transmission protocol.
I think Tidal (and most of the others) use RTSP to stream which should be buffered by the playback device. It is essentially not a pure real time UDP stream but not yet TCP stream either.
I just have to add that even if a service used UDP it would not matter. When a packet enters the end users household, the damage is either done or it isn’t. The last meter of hifi ethernet will not have any effect. UDP in home network is really reliable unless there is something really wrong/broken about the home network.
If the digital signal is degraded to the point where you have unrecoverable errors, then the audible effects are not at all subtle.
But people here aren’t talking about ear-piercing screeches or audio dropouts.
The advocates have, therefore, shifted to arguing that RF noise from the digital side somehow percolates through your DAC and affects the analogue side of the chain. This is something that can be measured. But it’s clear that measurements are not going to convince anyone.
Nor does it seem to give anyone pause that the more expensive the gear, the worse it performs (or, at least is claimed to perform), vis-a-vis noise rejection.
Rather than comparing opinion. Can the cable believers record the output of their dacs with cheap cables and then again with expensive cables and the do the old invert phase so we can hear the differences that they are hearing please.
I personally do not believe that cables make a difference but this would be a very very very simple way to prove me and anyone else wrong.
I do wonder why someone who happily believes that their cables in their system improves the sound to their ears, would be even remotely concerned about the fact that there are others who don’t believe them?
Believers might reasonably turn it around and suggest that if you’re troubled by their claims, then why don’t you put the effort in to do this test, buy an expensive cable (on a 30-day return policy) and perform the experiment yourself to prove them wrong…?
Just be careful on quality of the analogue input device used to sample the DAC output. If there is no difference, some might suggest the input lacks resolution. if there is a difference, some might suggest that’s simply electrical noise on the analogue recording device. So I suspect it may not be so very very very simple…
Personally I am very sceptical as to whether boutique ethernet cables make any difference (except to the wallet). If they do make any difference, I believe that it is likely through the modulation/reduction in electrical noise and not as a result as having any impact on the quality or reliability of the actual data (it’s all 1’s and 0’s folks!).
Most of us who have experimented with cables and switches are quite happy with what we’ve settled on and not in the least interested on spending time proving anything to anyone.
@Bill_Turner’s is. More generally, I’d be curious to know: of those of you who say you hear a difference between different ethernet cables and/or switches, how many of you are using NOS DACs?
Some of the believers ehm… believe that the content of streamed audio is altered during data transport over the network before it’s processed by a DAC or by player software. The alteration they say is due to electrical noise and undefined influences like “phase shift” whatever that may be in networking terms.
So there are people who are absolutely convinced their music software receives and transmits bit perfect audio files and yet these bit perfect audio files are tainted in some way by electromagnetic interference.
They just overlook the glaring contradiction in this belief.
You probably misunderstand them: the contention is in almost all cases not to do with the data being corrupted though some claim that intensive error correction can impact SQ. Generally the view is that the cables/switches have an impact on the performance of connected analogue systems that impacts the SQ
You seem to ignore the fact that the signal transported by ethernet or USB is a modulated analogue voltage signal which can be affected by EMI/RMI.
Only the interpretation of the signal is digital information.
Even if a few bits per second are altered we won’t hear any difference because error correction/interpolation will even it out.
If you have (noisewise) a clean surrounding in your Hifi set up, you won’t hear differences, but in set ups with a lot of factors that disturb/interfere with the signal there can be audible effects (still subtle) when exchanging cheapies with better cables.
There is no clean YES or NO, but a huge „it depends“ (on your system and its surrounding conditions, on your hearing capabilities).
I have just upgraded to a AudioQuest Diamond Cat 7 cable. My network data was indeed being corrupted, as I just looked at my bank account and the balance has been corrected by -10.000 euro.
All digital signals are represented by analog quantities (voltage, current, etc.) in the real world. This is no different whether the digital data represents an audio signal or anything else. If EMI (or RMI - what is that? I guess you meant RFI) causes the data to be transferred incorrectly this can be detected at the receiving end.
This is not true for Ethernet or USB. Each packet of data, containing multiple audio samples, has a CRC that the hardware uses to detect packets that have been damaged in transit. If the CRC is bad, the whole packet is discarded.
In network-based streaming protocols, the lost packet is usually re-transmitted in which case the integrity of the audio stream is completely unaffected. You will not get a situation where “a few bits per second are altered” in the stream of audio samples sent to the DAC.
In USB audio the dropped packet is not re-sent, which usually causes an easily-audible glitch in the playback - not at all subtle, and an obvious indicator that something is broken.
With the acclaimed Melco N1 series, we eliminated the data switch from the data path (server to player) and lifted sound quality hugely. But it was still clear that the rest of the network was having an effect. It was, perhaps, less obvious on lesser systems, but as soon as we fixed everything else, it remained a nagging problem.
Work started to analyse exactly what was going on, and the resulting S100 resolves those final nagging sound quality issues. Crucially, while the S100 is clearly befit for Melco owners who use the PLAYER Ethernet port, the difference an S100 can bring to non-Melco systems is quite surprising, especially for the heavy traffic created in a Roon environment, as well as for streaming services such as TIDAL and Qobuz, where the data has been seriously compromised on its journey from the distant servers into the user’s home. The S100 restores the magic of the source stream as no IT device can possibly do.
Needless to say there is no further information about precisely how the data from TIDAL and Qobuz has been damaged on its journey, nor on what “magic” is being restored.
Yes, or it travels 6000 miles with no issue, then gets completely corrupted to the point of sounding terrible in the last meter of cable from my data switch?
