Posts: 278
Threads: 7
Joined: Jun 2017
Reputation:
19
This isn’t music reproduction. This is simple file transfer. Not an audio stream. Not analogue sound. Not anything other than a transfer of a digital file over a digital medium. We aren’t sending this audio to Phantom over auxiliary cables.
Sent from my iPhone using Tapatalk
Devialet Phantom (White)
Posts: 2,443
Threads: 44
Joined: May 2014
Reputation:
84
This had potential to be a more interesting thread than most on the topic because the cables in question weren't physically connected to the Phantoms, but nevertheless it seems to have followed pretty much the usual course...a shame.
Roon (Mac Mini), Wilson Benesch Full Circle, Expert 1000 Pro CI, Kaiser Chiara
Warwickshire, UK
Posts: 39
Threads: 2
Joined: Dec 2014
Reputation:
0
02-Apr-2018, 09:09
(This post was last modified: 02-Apr-2018, 09:11 by alphafox.)
Interesting thread! It is quite funny how digital transfer is not understood by most people. Coming from the analog world, one may consider all signal transport as real-time and definitive, i.e. if one equipment tempers the signal, there's no way you can correct it later. Digital transfer introduces the possibility of asynchronous protocols, the speed of transmission being much higher than the required data throughput.
When investigating cable influence on sound quality, you should just wonder "is the digital signal synchronous or asynchronous?". In the case of early USB DAC connection, the PC and DAC were synchronous on the PC clock: poor quality of the USB cable could cause the loss of some bits, impacting the sound quality (not mentioning the poor PC clock quality introducing jitter...). However, with Ethernet connection (or WiFi), the transferred audio data is buffered and lost packets are re-resent asynchronously from the playback. So, the Ethernet cable cannot impact the digital data stream.
The only reason one may invest in high-quality cables is for electronic perturbation, assuming the DAC is of poor conception, not isolating properly the inputs... When it comes to Deviatet, I assume its engineers are competent enough to ensure the isolation of the input stage!
Synology DS414 > Roon Optimized Core Kit on NUC7i7BNH > Devialet Expert 220 Pro CI (upgraded from D120) > Cabasse Galion IV (vintage)
Posts: 1,151
Threads: 14
Joined: May 2014
Reputation:
26
02-Apr-2018, 09:24
(This post was last modified: 02-Apr-2018, 09:33 by Rufus McDufus.)
Although streaming audio over ethernet is typically done using UDP which is better for real-time transmission but does not retransmit. I believe from someone's analysis of Devialet AIR on this forum a while back (apologies I've forgotten who it was) it looks like the Devialet AIR protocol does have some sort of inbuilt error checking (and retransmit? edit - no, it couldn't could it?) to make it behave a little more like TCP.
Ethernet NICs do have (or should have) galvanic isolation built in, which I believe is effective against common mode noise but doesn't protect against all forms of noise. I suspect Devialet would have put a standard off-the-shelf NIC in their amps.