I was including/assuming bit perfect and galvanically isolated. My impression is most here would consider that a requirement. I guess the thought or question I have about that is, is it not possible to get bit perfect without galvanic isolation? So then you add G.A. and now it sounds better/different. Still bit perfect…
So I can imagine that it is also possible that Roon sounds different, even when BP and GA’d, just as many swear by the sound of Audirvana or whathaveyou, as they feel it sounds better. I may be wrong, but get the impression a number of folks use other players with Roon.
Bit perfect transmission and galvanic isolation actually address separate issues.
In my system, even with pretty noisy ethernet, it’s easy to get bit perfect transmission. Different computers, routers, cables, and software, it’s all bit perfect. I assume this holds true for most of us - moving bits is now mature technology.
Those of us with DSDacs have jitter taken care of for us as well, with the Dac re-clocking the bits without reference to prior clocks -so what’s left to affect SQ? Noise.
Galvanic isolation blocks the transmission of noise from everything upstream. Theoretically, it should equalize every bit perfect bitstream up until that point. It becomes difficult to find a reason that different bit perfect streaming software would sound different under these conditions.
Note that this type of comparison just applies in the specific situation described. It does not apply to something like HQPlayer which may improve SQ by doing something better than the Dac, but it’s not bit perfect (and not intended be). And it doesn’t apply to something like the Aries G series where the bitstream downstream of the ethernet conversion is markedly different.
My original response was in response to the implication that Roon negatively affects sound quality. In my system, introducing galvanic isolation, made differences due to streaming software difficult to hear. This (unusually) seems to agree with how things should work theoretically.
The sonictransporter i5 is a great device at almost half the cost of the Nucleus. The Nucleus servers are over priced for what they do. The i5 works great for hosting the Roon Core.
Paul and everyone, any updates on the server and Octave? I know there has been a huge amount going on and a flood of killer products… but this is one I am really waiting for! TIA
Me too! We’re moving steadily along. The schedule, at the moment, is for an alpha version of hardware and software to be completed and begin testing in late October. Given its hopeful success that would mean we could potentially look for a limited beta test
phase in December and launch of the product early 2020. I’ll keep you in the loop when these things happen.
Just by interest…assumed a different development effort and parts cost…what’s dominating the cost of components like the DS, DMP, Octave Server, BHK pre as they have an identical price? Is it the chassis?
PS uses the same margins that most other high-end companies use. I wrote about their approach in Positive Feedback a few years ago. No idea how Apple prices things.
Will the Server have a I2S input that would allow a stream from a DMP or PerfectWave Transport to use the new hardware interface for better sound quality or import the data/file to use on the sever? @Paul
That’s what I mean. Given those 4 mentioned products and adding the Regenerator with the same price and chassis, I’d say the DS and Octave will have by far the highest development effort while e.g. a smaller regenerator is a spin off of the bigger ones with overseeable dev costs. Parts of all those products will be not that differently priced and chassis is the same anyway. So to me it seems development costs are more or less unconsidered and the price represents chassis and parts cost mainly.
Not sure if we can see this independence of development effort elsewhere in other industry segments and if we see this assumed difference in development effort within an equally priced component series elsewhere in the audio business.
The Octave server will certainly have I2S outputs to feed DACs, though if you’re asking about inputs to use DMP as a ripper, then no. It’s likely we’ll build a ripper into the server.
TIA