Higher End DACs

RF isolation is needed to determine DAC differences. That is, RF noise in signal, power or radiated thru free air affects how a DAC sounds. 
Regards xtreme servers ...well if the output is optical and the server is 100% RF isolated from the DAC, they sound the same as a $100 raspberry Pi. Exact same.
Im in the Toronto area and able to plainly demonstrate this.

@dmance,

Thanks for chiming in...some of us are well aware of the benefits of isolating RF noise from the streamer. It has been discussed in the following thread to an extent, 

https://forum.audiogon.com/discussions/network-switches

Feel free to add your feedback there or start a new thread. I think this very topic of noise isolation deserves a dedicated thread and we can possibly learn and educate ourselves from your experience.

@lalitk
Thanks. I will (gingerly) post on that thread. I am going to turn this industry on its head so i need to be careful. The issue of RF noise (basically any EM energy above the audio band) is grossly under-appreciated as being detrimental to ultimate sound from a DAC. There is so much misinformation about tweaks to the upstream digital chain. Servers, cables, power supplies, switches, reclockers, etc. - all do nothing to the fidelity of the digital stream to the DAC. It’s always 100% perfect. The bugaboo of jitter is a 1990’s issue - modern DACs are impervious with double-buffers and separately clocked outputs.
It may be incredulous to many (most) that digital electronics radiate RF noise across meters of open air or via galvanic signal paths or AC mains to enter a DAC to affect the final D/A stage through perturbations in clocking or reference voltages. So all those tweaks ...all they do is change the radiated emissions profile; modulating the RF correlation to become manifest as subtle deviations in a DACs output waveform timing and amplitude: hence staging, timbre, detail.
I wanted a solution not a partial fix so I solved this for myself (and its a commercial Audiowise product) by putting my entire digital chain inside an RF isolation chamber (a 90dB attenuation Faraday cage) with only optical allowed in/out. Bingo - zero audible difference between a Raspberry Pi, an Intel NUC or an ’audiophile’ server. No difference between a $5 USB cable or a $500 one. No difference between using a factory switch-mode power supply or a $1500 LPS. When all these tweaks are contained inside a RF chamber, no RF noise escapes and the DAC always sounds its best (as it was designed).

Dmance your findings seem strange, first you don't list any of the components you are testing. 

Second you have only optical going in or out is that an optical cable from the dac to the server? Or some ethernet to optical and optical back to ethernet device?

Also why would you expect there to be an audible difference between a Rasbery Pi or a Nuc, both of these should sound the same. The issue is which is the audiophile server that you used and how was it compared? 

Just because you deem something to be a "audiophile server" doesn't mean it really is, the exact test conditions of what you used and how you used it may shed some light that a: you are 100% correct and it is all RFI/EMI that we are hearing as the cultprits in helping create different sounds out of a dac or

b. Your testing methodology of conclusions are wrong, what was the level of the listening gear being used to test the sonic differences? If so how were you able to then 100% shield any output cable from allowing back in emi/rfi and therefore invalidating your findings?

Just because you have access to a really big faraday cage doesn't mean your findings are correct.

Just food for thought.

Dave and Troy
Audio Doctor NJ

@audiotroy 

I have Chord Hugo TT2 to headphones (or direct to Voxativs) isolated from AC noise on a battery and Toslink signal input. With such a configuration, nothing upstream should make a sonic difference, right?  The TT2 is impervious to input jitter, so why does a server make a difference? Please don't claim that that bits are 'better' in any way ...yet in open air any upstream changes can sound different, why?  If a switch or power supply or cable changes the sound but the bits are the same going into the DAC, what else could account for this? Hint: its all about RF noise.

My test server was a Baetis (circa 2016), my NUC is a 5i5RYK, my Pi has a HifiBerry Digi+ module.  My test equipment is my ears combined with a SignalHound BB60c to correlate measured RF noise in the signal and proximate to the DAC.