I don’t know the internals of the Rossini Transport well enough to know for a fact whether it’s actually better to upsample on it rather than on the Rossini DAC.
However, on the basis of the picture I’ve seen from Stereophile Review of it, the Rossini DAC’s Control Board - where all of the main digital signal processing happens - appears significantly more powerful than that on the Rossini Transport.
Although as Phil indicated, there’s additional processing going on within the DAC to feed to D-to-A conversion stage, but you’re literally wasting significant processing power on the DAC by upsampling on the transport.
With the Vivaldi stack, it very clear that upsampling on the separate Vivaldi Upsampler is significantly better than on the Transport.
In any case, stick with whichever is your sonic preference, just don’t do it on the basis of imagining there’s “overload” going on.
If the transport is feeding data to an internal buffer inside of the DAC, and then the signal is further processed inside the DAC, why would this be the case?
I see that an external clock could provide timing benefits itself, relative to an inferior internal clock, and if there are multiple components in the chain, i.e. Vivaldi Upsampler to DAC, but there appears to be disagreement on the prior point, according to MSB’s and Wadax’s architectural choices, which instead isolate the power supply, but co-locate the clock with the DAC.
So this weekend I shall do a critical
listening by running the Transport in pcm mode connecting only one AES output to the dac.
But wondering if the 1st stage (as clarified by @Phil ) of upsampling in Transport is not as good as or better than the Dac why someone will pay for $29 K for the transport ?
@phil one quick qs on ‘whichever transport method you use between the Transport and the DAC (whether AES or S/PDIF) the DAC doesn’t have to extract / create the audio data timings from the source audio data’.
As far as I understand spdif data contains the clock information. So are you saying if external clock is used the dac will ignore the clock information from spdif data and use the clock information from the external clock ?
This is not at all as paradoxical as it sounds, @T38.45.
IIRC the actual crystal oscillators (the clocks themselves), one for each “base frequency”, are the same in the Vivaldi DAC, the Rossini DAC and the Rossini and Vivaldi clocks. The differences (aside the casework ) are in the functionality and the supporting circuitry.
To make the absolutely most accurate, no expense spared, word clock you would want to reference it to a highly stable reference clock. But that requires connectors and supporting circuitry so as you try to make things cheaper it’s one of the things you can leave out.
Consider this:
Vivaldi Clock: 4x2=8 word clock outputs (allows connection to four units at switchable sampling frequencies), one reference clock input
Rossini Clock: 1x2+1=3 word clock outputs (allows connection to one unit at switchable sampling frequencies and one unit at a fixed sampling frequency), no reference clock input
Lina Clock: 1x2=2 word clock outputs (allows connection to one unit at switchable sampling frequencies), no reference clock input
They each support the full functionality of the product lines that they are part of:
Vivaldi Clock supports connection to Vivaldi DAC, Vivaldi Upsampler and Vivaldi Transport (all switchable as Vivaldi Transport supports SACD as well as CD)
Rossini Clock supports connection to Rossini DAC (switchable) and Rossini Transport (non-switchable as it doesn’t support SACD)
Lina Clock supports connection to Lina DAC (switchable)
So in actual fact it is no more surprising that the top-of-the-range word clock has the ability to connect an external reference clock than that the cheaper siblings have rationalised away that ability.
I have only one target for a clock- that’s my Rossini DAC.
So the benefit adding a Vivaldi clock (as I understand you it has more output clock connections) doesn’t count for me. Ok, the Vivaldi clock may be more accurate, but does one really hear that?
For that money, I would go for a Vivaldi DAC first and add a Rossini clock to it.
add on: that’s MY guess…I’m 100% ok if somebody say it’s better the other way around
I have not compared all the permutations but others have (well maybe not all but enough to be able to draw some conclusions and vote with their wallets).
Afaik the price difference between Vivaldi and Rossini DACs is greater than that between the clocks so the sound quality improvements have to be judged relative to the respective price tags. But I would still say it is worth auditioning (if possible) rather than just assuming based on some theoretical rationale.
It is definitely the right strategy, the Vivaldi Dac will make a much bigger improvement than a clock. You don’t even need the Rossini clock…But you will need a streamer…The Vivaldi Upsampler is the best choice, but rather expensive.
You buy a second hand dCS Network bridge, you save money until you can afford the Vivaldi Upsampler…
After a 4 days of home demo of the Rosinni Clock, I ordered one this week.
Thanks @Phil for suggesting the home demo. And my dealer Music Lover Audio was also prudent to provide me a demo unit within few days.
Thanks to others here too for your viewpoints in this regard.
With clock in my system, I found that the sound is more lush (little less punchy). May be because of the fact that my rest of the system is all tube. But overall very engaging sound. And probably the CD playing through Rossini transport is more benefited with the clock than the network music from streaming service/music servers.
I am sure there are benefits to adding the Rossini clock and congratulations on the purchase. I would like to explore a broader issue here, re: dCS’s architecture on this product.
The stated reason for having an external clock is to control the timing of multiple products that form a cohesive signal path, for example, the Vivaldi DAC and the Vivaldi upsampler. This seems logical and compelling.
However, I do not understand why, for the Rossini (which I own), it would not be better to take a different approach and (a) separate the power supply from the DAC, thereby reducing noise inside the chasis; and (b) simply improving the internal clock.
If one is not using a transport, there are no separate chases to coordinate (unlike the Vivaldi), since the upsampling and D2AC happens inside the same box.
And with that comment/question, I welcome your thoughts, and I prepare to be flogged… ; )
Every company choses architectures on the basis of their expertise and history. dCS has, from almost the very earliest days, been experts at external clocking. Continuing to build on their strengths makes absolute sense, why would they do otherwise?
That said, at the end of the day, what really matters is the final results.
On that basis, objectively speaking, dCS’s systems generally outperform (or at a minimum, equal) corresponding systems which focus on internal clocks only and/or better external power - MSB, EMM, etc etc etc. - just look at any Stereophile measurements.
I even remember someone from dCS saying that the clocking units themselves were the same between the DAC and externally-housed versions. (Don’t remember which ones. Will try and find the thread.)
That feels like some way towards saying “it’s better because it’s external”, though I suppose there may be ways to get isolation to the same degree with multiple PSUs, larger cases etc.
All of these design decisions must come with trade-offs. One of the bummers for the dCS route is added cabling. An umbilical power cord is simple enough. Eleventythree clock cables less so.
I have borrowed a Rossini clock on two occasions for a few days each. I use a Melco wired to my Rossini Apex via Ethernet. I only listen to files on the Melco. For the life of me, I could not hear a difference with the clock. Clearly, I’m in the minority.
?
