Question for the community – am moving from Naim and would like to understand dCS nomenclature, system architecture.
On home demo, have found that a loaner Rossini + Clock sound terrific when fed via SPDIF (from a Naim UnitiCore) but rather disappointing when used with Tidal either via the dCS Mosaic app or from a networked Roon Nucleus +.
Will be shortly auditioning the Vivaldi Apex. In dCS nomenclature, I believe that the Vivaldi Upsampler adds what others may call ‘streaming’ functionality is this correct?
Does it lift the Vivaldi DAC’s sound quality or is it mainly a matter of providing a network / Roon / streaming interface, and adding control by the Mosaic app?
If I hypothesize feeding the Vivaldi DAC either via USB or SPDIF (from a high end server), then it seems that any additional outlay would be best allocated to the Vivaldi Master Clock rather than the Upsampler – is this the general consensus?
…and I may just be your ideal person to help. Given my prior twelve years at that establishment before joining dCS!
OK - That’s interesting … I don’t have my Rossini and Clock here any longer as they were needed but I don’t ever remember feeling that there was any kind of downside to using TIDAL via Mosaic or Roon either … both are definitely bit-perfect transports too.
Sort of … One function is definitely streaming but the other is the upsampler functionality.
Now, if you had the incredible pleasure of conversing with me in my days when I was still “there” you hopefully found me quite straight talking and not at all happy with fluffy HiFi nonsense?
So, upsampling annoys me, it annoys me because logical me says it can’t work and I’m not going to insult you by referring you to a dCS white paper telling you what it does and how amazing it is but it IS a very nice tool to have in your audio toolbox and there’s a lot of independent information out there that goes into what it does as far as performance in the real world and I’m sure many forum members who will tell you here what it does and that it’s a major dCS feature.
It is and you’ll be able to try it for yourself when you demo the Vivaldi Upsampler…
It definitely provides that streaming and network functionality but IMO the upsampling is the main benefit…
The Vivaldi Clock is of much more use with the DAC and Upsampler (and transport if you have one) as it gives a master reference clock for the devices to sync to … it doesn’t have the same functionality with external sources that aren’t synchronised to the Vivaldi Clock.
In fact, if your external unsynchronised sources drift too much then you can end up having to turn off using the Vivaldi Clock as a clock source on that input and use the input signal as a clock source instead.
Hopefully that answers your immediate questions … of course if you have more then please drop a reply and I’ll try to help as much as I can.
And may I chime in as well as a current Naim 552/500 owner who replaced his Linn Klimax DS with a Rossini+Clock 2 years ago:
dCS recommend feeding their DAC’s via ethernet as this will produce the best sound. Bypass the Naim Core - no need for it and many do not like the way it proprietarily rips and locks down your music files.
Ditch Tidal and move to Qobuz. Many Tidal tracks are MQA while Qobuz tracks are straight CD quality and better. Their hires are true and not MQA “enhanced”. Enough said.
My Rossini fed by an ethernet cable from my Cisco 2960 router and connected to my 552 via Chord Music RCA>DIN thoroughly walks over my Linn KDS/3
Do you have a view of the clock drift tolerance that the Mutec MC-3+ supports on their synchronous (AES3/SPDIF) inputs?
They talk about it, but there’s zero technical details or specifications. The do provide the internal clock accuracy/jitter specification (+/- 0.1ppm / <1ps (RMS), but thats not the tolerance levels for the inputs.
Or has anyone measured the Mutec’s actual performance?
If you feel that that gives an improvement that justifies the expense then fine but otherwise it’s another piece of kit that is being introduced to resolve an issue that doesn’t necessarily need to be resolved … using the source as clock for that input will do that too.
@Phil - of course I remember: years ago when I lived in Canada I remember you remotely logging into my HDX and you’ve helped subsequently since I relocated to the UK - thank you!
The loaner Rossini + Clock have gone now but it was marked that they sounded better via SPDIF than using the Mosiac app for Tidal or using Roon. I do have a Melco S-100 in the loop - could this be relevant?
Re Vivaldi Upsampler - the Ring DAC White paper indicates that all incoming material is ‘oversampled’ and converted to 5 bits. So does the outboard Vivaldi Upsampler do some of the oversampling / upsampling which would have been done by the Vivaldi DAC, thereby reducing the processing load on the DAC?
Also, the Mosaic app had setting for DXD, DSD DSDx2 (from memory; with filters for each). I understand the filters but not the first three options - what exactly do these settings refer to? The DAC knows whether it is receiving PCM or DSD, and oversamples / converts to 5 bits anyway for all material, so what are these settings for?
Re Vivaldi clock - no benefit if you use a Vivaldi DAC with an external (asynchronous) server (eg Taiko) or library (eg Melco N10/2) but for DAC used with Upsampler, the external Clock’s benefit is in sync’ing the DAC and the Upsampler - is that right?
Ah - ok - reading through the product manuals has clarified, together with a realisation that the terminology is a little confusing. It seems that the Vivaldi DAC takes the incoming PCM or DSD and oversamples it to a high frequency 5-bit datastream going into the Ring DAC. If you add the Upsampler, then the PCM or DSD is upsampled to either high-rate PCM (ie DXD) or DSD, selected by the buttons on the Mosaic app, before going to the Vivaldi DAC. The Vivaldi Clock serves to sync the Upsampler and the DAC (and transport if present) to a higher quality clock signal. So, with regards to the filter options, at what stage are these applied? And does the community think that the reason why SPDIF input into the Rossini + Clock sounded better than Tidal via Mosaic or Roon (Tidal and Qobuz) might have something to do with the Melco network switch being in the loop? Unfortunately, with the loaner Rossini + Clock having gone now, I can’t further experiment.
I am happy for anyone who, after trying an “Audiophile” Ethernet Cable, “Audiophile” Ethernet Switch or “Audiophile” Ethernet Reclocker within the context of their own system, feels that any perceived change is worth the expense and buys such a unit …
I will also say that everyone should be aware that it is very easy to “tweak” yourself into having an unreliable network after starting out with a reliable one … I did a presentation on exactly this a few years ago for Clarity up in London (which was quite terrifying).
Certainly when trying to diagnose many customer issues one of the very first steps that I would want to do is remove any Audiophile network tweakery (including audiophile Ethernet cables) …
You are right on both counts, but there is a key difference between the Rossini and Vivaldi systems. What the Upsampling setting in the Rossini DAC does is selects how the data is oversampled, before being converted to the Ring DAC 5-bit format. Taking incoming PCM data and upping it to either DXD or DSD before being converted to the Ring DAC format does actually cause a difference in sound between the two options. There is no right or wrong there, it all comes down to what the listener prefers, which is why we offer the choice.
This is not presented as an option in the Vivaldi DAC as in a Vivaldi system, the Upsampler is assumed to be handling this. You are right, separating it out into a different box does indeed cause an increase in sound quality, for a variety of reasons (gentler processing, less micro-drains on the DACs PSU, that type of thing). It actually came as a surprise to us the first time we tried it in a separate box and heard the difference!
Thanks @James , @Phil - much clearer now. One question (for the time being) - whatever the listener’s choice of filter, is this filtering done between the Upsampling and the Oversampling / Modulating steps, or after the latter…?
I guess it depends on what you define as tweaks. I inserted a switch with a optical connection that ends about 40’ away and connects to a optical module that coverts back to Shumyata Ethernet for a short 1 meter connection to my Upsampler. My DCS/Roon set-up is rock solid.
It happens during the Oversampling stage. Upping the ‘sample rate’ of the signal usually involves inserting samples with an amplitude of 0 between the actual audio samples, then filtering it digitally - so this all happens as one process.
Having tried myself a melco S100 on my rossini/clock, it didn’t last long, and it made things worse for me and i believe a few others have also had the same thing.
I don’t run any so called special switch and just a cheap cisco switch to give me the required amount of ports and it sounds lovely streaming as well as playing storted stuff from the melco N10
