This couldn’t be further from the truth. You got your analog and digital mixed up. Remember, the DAC is an analog device.
I speak about this in this topic about Ethernet cables:
We are saying the same thing. True DAC is analogue but before all it is digital. The comming in signal is digital. The treated signal is digital and the output of the DAC is analogue. But we can discuss this all day. The point is that the quality of the electrical signal will greatly affect the quality of the system outcome . My 2 cent
On the ANALOG side, “special” power supplies do not/cannot alter the digital signal. Otherwise, I’d want the purest PSU I can find before doing any online banking. But we can also discuss this all day 
Does that refer to the ASIO drivers? They are compatible with ROCK? This interests me (a little) because my amp apparently will accept DSD256 with Wintel/ASIO, but only DSD128 otherwise. Not that I’d switch platforms for this reason alone.
Not sure on the DSD front. I just know that the win 10 drivers for my DAC are problematic and involve lots of hacks to get the unsigned drivers past windows security. I had similar problems with another DAC a few years ago. With ROCK it was plug and play.
I use a linear power supply for the roon server, and the ssd storage. All digital.
For the analogue part, I use a power filter.
I have tried many different ways, and this works best for me.
As a complement, an interesting conversation not only about Ethernet, but the “digital” part in audio in general with three experienced developers who should know what they are talking about:
There’s no such thing as digital
1 Like
This is an interesting article and the comments (which were even more interesting) suggest no one was persuaded one way or the other. I’m not technically proficient enough to evaluate the merits of the article, but it does appear to be a significant amount of hand waving. Having said all that, I do have a linear PSU in my gear mainly because I hate clunky wall warts. It’s just a cleaner approach to providing power to my various components. I’ve never heard an audible difference one way or the other.
1 Like
More to the point, it’s from 7 years ago. Technology moves on.
1 Like
The technology perhaps, but not the physical principles.
It’s almost total nonsense. As most of the comments on the page point out.
2 Likes
And what are your qualifications to judge that? It should be quite easy for @danny to ask Steve Silberman for background informations 
People, why don’t you just keep it simple!
Do what works for you and your ears.
Linear or not, digital or analog, it does not matter as long as you are satisfied or notice a difference in your system.
At my place, the general current is relatively clean especially at late hours, and during the day I do notice a difference with and without my power tweaks. it works for me, and I am satisfied.
I am personally not looking to understand or prove the physic behind it, although some people might be and I do respect but the answer or truth won’t make a difference to my hearing 
1 Like
Gordon is mostly talking about cases of digital with no error correction and also about the conversion, an analog process.
Charlie pretty much is in agreement that digital can’t be screwed up by analog things.
Steve’s one comment is just flat out nonsense. Digital bits are not square waves suspectible to modifications. If they were, nothing would work in the world. Digital bits do not just go bad. Everything in computing hardware, software, and design is made to prevent that. This is the part where information theory trump’s electronics.
6 Likes
And it’s really important to emphasize that digital audio is nowhere near being a “challenge” in that regard.
- Consumer-grade (i.e., dirt-cheap) ethernet operates at 1 Gbps.
- USB 3.0 operates at 5 Gbps.
- “High-res” digital audio is, at most, a few 10s of Mbps. (“CD-quality” audio is more like 2 Mbps.)
In other words, digital audio represents less than 1% of the digital channel’s capacity.
You don’t need a fancy ethernet switch, or “audiophile” digital cables. Ordinary consumer-grade hardware is designed to handle 100 times that much that much data, flawlessly error-free.
6 Likes
It’s not about the data but electrical noise going down the cables along with the data. But please, don’t do any power or device tweaks to your network as I’m sure whatever you’re running was just perfect out of the box. But my network has been tweaked considerably (Cisco 2960 switches, optical modules, LPS’s and BJC cables) and it has paid off in sound I never imagined with my level of analog gear. Even if each item brings only a small percentage of change to the sound, those all add up to a fair percentage of the sound quality. But again, until you actually do it and hear it, go ahead and keep blowing it off. Your loss and my bliss.
Well I have relevant qualifications and experience but frankly that’s all bit unseemly, and also largely irrelevant on an anonymous internet forum. There is no point in a dick waving contest. Besides reading that my bullshit-o-meter was in the red pretty quickly. If you want to take that kind of stuff on board good luck to you. I’ll pass.
4 Likes
And how is that relevant to anything?
Do you have a mechanism by which “electrical noise” on (say) the ethernet cable ends up in the final analogue output of the DAC?
A far more direct source of would-be noise is the power supply of the streamer or DAC.
That myth has been tested extensively. See, e.g. here and here for tests of a streamer and/or DAC powered by a) a switching power supply, b) a linear power supply or c) a lithium battery.
If the power supply of the streamer and/or DAC produces no detectable difference in the analogue output, there’s even less reason to believe that the ethernet switch could possibly make a detectable difference.
2 Likes
@Jacques_Distler. There are quite a few examples of rf noise being problematic and something to avoid in ethernet and other audio cables specifically because it can affect the analog output.
It’s why Rob Watts recommends ferrites to be placed on BNC cables for Chord Dave/Blu. Here’s the thread if you want to read it: https://www.head-fi.org/threads/chord-electronics-dave.766517/page-711
It’s why both Lumin and Sonore are now putting optical network connectors in their streamers. I would expect these to be standard fare in most if not all high-end streamers in the near future specifically to prevent this.
Whether you have a rf problem in your setup depends on your setup and of course that affects whether you have an issue, but the problem does exist.
Could go on, but you can find examples of this from reputable sources if you search.
I thought the same 5 years ago about ethernet connections. But, alas, no, not standard fare.