This is exactly true, in this context bits are perfect, and they have to be.
By reading your comment above, it seems you refer DAC as part of the analog stages. Then I read again your other comment
I think we are probably talking about the same concepts The gap is probably interpretations of terminology. To some of us, or at least myself, see the digital input interfaces of a DAC as digital, the analog output interfaces as analog. Perhaps you see it the same way, too. The way I see it, the DAC chips (or r2r circuits) would be the demarcation points, and just the way I see it.
I truly appreciate that you took and shared the observations. What is your choice of connectivity between your PC and DAC, and what you may have done or plan to do for SQ? Asked also for the benefits of @richard_a.
This is an excellent thread. They say you never stop learning, how true! I also apologise to the OP for muscling in.
Before I bought my Nucleus I started a topic re “Nucleus direct to a DAC”. Most members that replied to my post suggested that a ‘Streamer’ is not required and that any dealer that says otherwise is trying to rip you off! I’ve no idea to be honest?
I researched DACs and ended up buying a Matrix Audio Pro Sabre MQA DAC, and I’m happy with the SQ quality. However, updating the software is a pain in the butt! (5 updates in the past 10 months) It requires a Windows Computer (Didn’t know that at the time of purchase) to attach to, then the update goes through several stages. Cut a long story short, I’m a iMac user that can’t use Bootcamp because the iMac is too old. So here I am again, gone round in a circle, looking for a new DAC/Bridge/Streamer.
I love Roon and I love my Nucleus. (Which replaced my ageing NAS) I just want to get the best out of it, to complement my Naim 282/300DR/Kudos 606s.
Any new ideas would be welcome.
Server PC (win 10 LTSC & Roon)
JCAT NET Femto
25m Cat5e
Lumin U1 USB out
ifi micro USB3.0 with 9V LPS
Accuphase DC-950 DAC USB in
The main reason for USB is DSD native. I lived with Roon conversion DSD-> 176.4 for 12 months, but wasn’t the same resolve as native processing by the DAC.
For the PC, the chassis parts were bonded so that the 0V of the ATX PSU to the frame is 0V and not some millivolt. This change resulted in a direct increase in soundstage height and a filled image from right to left. My guess is the noise that’s present between the 0V from the HDD to mobo to expansion ports has gone back to the ATX PSU and avoided the Network port.
A similar change was made with another computer using the JCAT USB card with similar results, albeit with modest components from the PC to the DAC.
The DAC is not always, but most often the first part of the analog process. How could it be digital if it produces analog?
@R1200CL, in Nuno’s interview, while there is more “hand saving magic” than I like to hear, nothing I heard is in conflict with my statement about ethernet jitter. Is there something specific you’d like to point out?
The next person you should be dropping a link from is Gordon Rankin. I’ve also read a lot of Gordon’s stuff and spoken to the man too. He doesn’t go near this topic because he also understands digital abstraction. His discussions tend to stick around the topic of cleaning up USB error rates (which, unlike the protocols that sit on top of ethernet, has poor reliability) or noise transmission (no argument there, other than who’s responsibility it should be).
I totally agree, and hope Ethernet will be widely adopted as the preferred interface and medium for streaming.
Heck, why not make Roon a cloud service
We will probably benefit from industrial Ethernet, like moving towards fiber and 10GB.
At least we shouldn’t expect jitter, since it’s claimed high speed internet to be jitter free by the specifications. (Ref. stressed eye pattern test requirements).
Why I also suggested incorporating a SFP+ cage in Nucleus.
But who knows what future bring. Maybe IoT and 5G technology will affect how streaming is transferred locally from the ISP.
Followed the link and saw your write-up. I have not had seen similar work like you did, amazing! I plan to study you post in detail and see how it may be applied to my setup. Many thanks.
Thanks for tip about him. Bit errors is another issue. But you have strengthened my belief in avoiding USB if possible
It’s good you agree with John’s White paper. So hopefully we also agree.
The essence is something like this:
Jitter and phase-noise with Ethernet and USB packet data have nothing at all to do with changing the bits. It is all about how the low-frequency effects of phase-noise (and leakage currents) modulate the ground-plane (and chip power networks) inside the endpoint devices. Modulation of the ground-plane creates clock-threshold jitter—which effects subsequent chips and ultimately generates more of the same.
I responded to someone who gratuitously attacked the high end audio Industry in a thread related to Ethernet and USB cables. His post remained up my response was flagged as off topic. I see.
I’d be happy to suggest a backup scheme, but it would depend on where and what you currently have your music on. PM me if you want to have a discussion about that.
The gap is probably interpretations of terminology. To some of us, or at least myself, see the digital input interfaces of a DAC as digital, the analog output interfaces as analog. Perhaps you see it the same way, too. The way I see it, the DAC chips (or r2r circuits) would be the demarcation points, and just the way I see it.
