"Audiophile Grade" Ethernet Switches - The new generation

[thumb5]

...
That means there are more than one clocks in the chain (streamer + Dev?) and the Dev clock rules in the end, right?

Yes, that's what conventional audio engineering says: in a buffered DAC -- which a network DAC like the Devialet must be -- only the final clock matters.  The clock used by the Ethernet interface is fixed and completely independent of the audio sample rate: for example, if you're using a 100base-T (100 Mbps) Ethernet interface it's clocked at 125 MHz no matter what data it's carrying (audio, file download, printer output, ...).

The networked DAC must have a memory buffer to assemble and store the sequence of audio samples delivered to it in the Ethernet frames.  Any information about the timing of arrival of the Ethernet frames is lost by the time the audio samples go into in this buffer, because it is irrelevant to the rate at which the DAC has to get the samples out of the buffer and convert them to analog.  The DAC clocks samples out of that buffer at whatever audio sample rate has been agreed between the streamer and the DAC; it's the accuracy and stability of that final clock that affects the analog output of the DAC.

But if we add a better clock in the path it helps in therms of performance, correct me if i´m wrong, please, computer audio stills a field with a lot to learn, for me.

The Ethernet clock is independent of the audio sample rate, and is not used by the DAC to generate the analog output (because of the buffering), so any changes to the Ethernet clock should not make any objective difference to what comes out of the DAC.

Those who make audio-grade switches will argue differently of course, but as far as I know there is no compelling (repeatable, objective, non-anecdotal) evidence that putting "better" clocks in the Ethernet path makes the analog output of the DAC more accurate.

There are potentially other reasons apart from clock accuracy and stability why using different Ethernet switches in the path might have some (arguably very small) effect on the analog output of the DAC, but those are also controversial and haven't been objectively demonstrated (again, as far as I know).

All that said, I wouldn't presume to tell you what you might hear if you try it, because it depends on much more than just the hardware you're using.