19-Oct-2019, 11:55
The only thing that matters in terms of clock is the clock of the last buffer.
So either you can imagine an electromagnetic way that other clocks are influencing that last clock (this is one theory for Ethernet which is electrically coupled and what fiber based solutions intend to avoid) or you have to imagine that the content of the buffer itself has an influence on the clock of that buffer, which is not impossible, but would be a very poor design IMHO.
Given the order of magnitude of the delay variation in a network (milliseconds to tens of them) compared to the jitter of even a bad clock, let’s say that I’m highly skeptical towards anything that is before the last network interface (by last I mean counting them from the source to the DAC).
Given how deceptive our brain can be, it would take proper ABX testing to scientifically determine the question.
My 2 cents,
Jean-Marie
So either you can imagine an electromagnetic way that other clocks are influencing that last clock (this is one theory for Ethernet which is electrically coupled and what fiber based solutions intend to avoid) or you have to imagine that the content of the buffer itself has an influence on the clock of that buffer, which is not impossible, but would be a very poor design IMHO.
Given the order of magnitude of the delay variation in a network (milliseconds to tens of them) compared to the jitter of even a bad clock, let’s say that I’m highly skeptical towards anything that is before the last network interface (by last I mean counting them from the source to the DAC).
Given how deceptive our brain can be, it would take proper ABX testing to scientifically determine the question.
My 2 cents,
Jean-Marie
MacBook Air M2 -> RAAT/Air -> WiFi -> PLC -> Ethernet -> Devialet 220pro with Core Infinity (upgraded from 120) -> AperturA Armonia
France
France
