Dcs upsampling options

Happy new year DCS fans, wishing you all an up-sampled year :wink:

I use 70% streaming and 30% NAS , I have these 2 options:

  1. Bridge + Scarlatti upsampler + Scarlatti clock + Vivaldi dac
  2. Vivaldi up-sampler and dac

How far is option 1 from the sound of option 2?


Hi Eylo and welcome. There is lots of knowledge and advice to be gained here on the forum but in this case I think that your question is straightforward.

Which selection has the superior sound quality is easy; Vivaldi was a significant upgrade from Scarlatti and v.2.0 of the Vivaldi firmware provided yet further distancing from Scarlatti in sound quality terms , not least given the new Map options for Vivaldi (and Rossini) which allow the processor to run at double speed.

However you also need to take account of what facilities the options provide and their interoperability.

Firstly none of the second generation wordclocks ( Puccini, Paganini and Scarlatti) are fully compatible with third generation products ( Bartok, Rossini, Vivaldi). This is because the earlier clocks only provide a single frequency output at a time. Vivaldi etc. require a dual wordclock input each at running 44.1 or 48 kHz or integer . This is to provide automatic switching when sample resolutions chnage - as they may do frequently when using streaming services. Without the dual output you would need to manually reset the wordclock output each time the incoming sample rate changes. So Scarlatti wordclock + Vivaldi DAC is not a great mix.

The subject of streaming services also relates to the Vivaldi Upsampler. Scarlatti Upsampler only does what it says in its name. Vivaldi Upsampler is also a network player providing, with its dCS Mosaic controller, replay of local files from NAS, Qobuz, Tidal, Spotify, Deezer and a vast selection of internet radio stations. No Bridge is required NB: the sound from the Vivaldi Upampler’s network player is somewhat better than the Network Bridge was ( yes I have owned both).

Given the option to purchase either of your options there is no contest - it’s a no brainer IMO- the two Vivaldi boxes and maybe go for the matching wordclock later if and when you can afford it.

Thank you for the detailed answer , I hope that you didn’t text & Party at the same time;).

Yes , I’m aware that the Vivaldi upsampler is also a bridge.

Given a tight budget I wanted to know how far option 1 sound wise from 2. Say 70% from it?

I am testing now option 1 at home and strangely enough I played DSD , high res and mp3 without playing with the clock settings, does it means that the bridge ignored the clock automatically in certain files?


1 Like

I don’t have any answer as those format names do not necessarily indicate the sampling rate. DSD and CDs are basically the same @ 44.1 kHz. However hi-res can be 44.1 base ( 88.2, 176.4) or 48 base (96, 192). MP3 can be 44.1 or 48 ( the latter typical for internet radio) though most are 44.1. So I cannot answer your question as it varies from program source to source.

However I would point out that playing a source with the wrong clock frequency will still “work” in the sense of there still being sound. However it will not provide the best sound quality and eventually will result in glitches because the clock and source frequency do not match and cannot be synchronised.

I certainly understand tight budgets as everything that I own has had to be saved for. However I cannot give a % value of X over Y . This subjective figure may vary from person to person and in respect of their supporting system and their personal values.Just accept that Vivaldi is better. Really :smiley:.

Thank you Eylo!

I too am in the same boat. First of all thanks for your question, which I am about to also post :slight_smile:

On your question about DSD. Mp3 and streaming Hi-rez, if they are multiples of 44.1 then yes DSD in 44.1 on Scarlatti works. Had that mp3 or Hi-rez been in 48 or multiples then DSD/48 is needed, however, most of the regular DSD are multiple of 44.1 so shouldn’t be an issue with the word clock set at 44.1 on Scarlatti.

I would like to add to your question. What is the significance of dual AES on Scarlatti upsampler? All the manuals state Scarlatti upsample does DSD, is it DSD/64 or upto DSD/128 (DSDx2)?

And Pete while I agree to your point full Vivaldi stack may be awesome and upsampler is computer possibly some hardware gain and software tuning may have improved in 2.0, however, is that change really $21K. I read in one of your posts mentioning dCS does every new version from ground up, however, software driven I seriously doubt it will always start from ground up but rather an improvement over existing.

Given the price to results where does NB+Scarlatti Upsampler+Scarlatti Word clock with Vivaldi DAC compare to full Vivaldi stack?

I heard from a previous owner that Scarlatti upsampler is 90% of Vivaldi upsampler? Appreciate if anyone can validate or share their feedback on this statement.

Ok , some updates

Per the owner, the Scarlatti gear has been modified at DCS to match the Vivaldi dac auto clocking, similar to the Vivaldi one , there is no need to switch between the clock frequency.

The result is really good , I can’t compare it to Vivaldi upsampler, but we did compare it to analog and I preferred the digital in this case.


It is DSD 64. The matching Scarlatti DAC could not process DSD128 as this was an innovation with the Vivaldi, Rossini etc, generation.

Why Dual AES? This is to provide compatibility with legacy dCS converters. It is also used for DSD from dCS SACD transports where dual AES is necessary both for legacy reasons and to accept the encrypted DSD signal from this source.

Interesting, could you please get from the owner what has changed. Only way I know the clock does change automatically is based on input from USB, which was available as an upgrade or on the + version.

Only other option I can imagine is via RS232, which is also available on Rossini 3rd output.

1 Like

Thank you Pete for the clarification, this explains. So DSDx2 there is no other option than Vivaldi Upsampler.

It is clearly not a sensible comparison as Scaraltti Upsampler offers only upsampling. Vivaldi upsampler is both an upsampler and a network player, Scarlatti offers precisely 0% of the latter performance :smiley:.

It can only be a very subjective figure anyway. Even if one accepts it then the missing 10% might be the important bit . Otherwise it is similar to saying that a Ford Mondeo has 90% of the performance of a Jaguar XE. It does but a Jaguar owner would tell you that is not the point.

Scarlatti and Paganini 4 box stacks sounded pretty similar (electrically virtually identical) - OK Scarlatti SACD transport was a lot better . I have owned both Paganini and Vivaldi at least as far as all except the transport. The older system was very fine so 90% of the Vivaldi performance may not be too optimistic a figure if restricted to just upsampling performance. However that misses the fact that every single (not just most) recording that I play on Vivaldi and which had been played on earlier dCS iterations reveals new insights. From that viewpoint maybe Scarlatti system offers only ( figure out of the air again) 50% of Vivaldi’s performance subjectively. Who really knows?

Ultimately only the individual and his (her) wallet can make the decision after a period of comparing the two.

He doesn’t know, he bought the Scarlatti already upgraded by DCS to match his Vivaldi dac. My guess that it is not only s/w. He is not using usb , only coax and aes.


He cannot connect USB from DAC to clock. Vivaldi doesn’t have firewire, so it may be a tweak via RS232.

No, he is using a coax cable between the clock , upsampler and dac to pass the clock . The data between the upsampler, dac & transport is via aes.
No usb or rs232 are used .

The original incarnation of Scarlatti had a USB input on Upsampler. This was fairly limited but allowed the connection of a PC with an installed player ( e.g. JRiver) . The USB input accepted PCM only <24/96 or <24/192 with the USB class 2 driver installed on the computer.

This was superseded later by moving the USB input to the system clock. This added DoP and provided automatic switching ( note USB only).

Although later samples of Scarlatti clock came fitted with this upgraded USB input it was also offered as a retrofit to earlier samples. It was a factory job as the rear panel needed to be changed. I suspect this may have been the actual upgrade cited.

The signal output from clock to Upsampler or DAC is via S/Pdif.

I have never heard of any upgrade to Scarlatti clock to provide dual wordclock outputs to match the later Vivaldi’s input requirements. Such a thing would at first sight appear to be outside of dCS’ normal business methods as once a product is discontinued no further development or upgrades are usually offered ( though technical support/servicing is still offered wherever possible) . So unless verified by e.g. dCS staff I would remain sceptical about this claim.


Hi Pete
See the photos related to the auto switching clock concept. Sorry about the cable mess :slight_smile:

Thanks Eyal but unfortunately I can’t see what is going on as the photo is a bit out of focus so I can’t read the port identifiers. The 2014 upgrade indicates an extension of the frequencies that wordclock supports rather than dual outputs though I think auto switching was supported on the USB interface. Unfortunately the Scarlatti wordclock manual that is available online is from 2013 so predates the upgrade and is no aid in this respect.

It does look from the photo of the Vivaldi DAC that only one of the wordclock in BNC ports are in use which would not allow auto switching other than USB. However , as I mention it is a bit difficult to see.

No need to apologise about the cable mess. It goes with the territory and mine is probably worse.

Hi Eyal,

I think the cables to the Scarlatti clock is missing, that is the important piece in this puzzle. Between DAC and upsampler isn’t an issue.

Also do you see the clock switching from 44.1 to 48 or appropriate sample rate? Reason for my asking was in the picture above 96KHz is upsampled to DSD. I did not see any documentation by dCS saying 48KHz multiples upsample to DSD/48 so it may be possible 48Hz multiples upsample to DSD, (equivalent of 24/176.4KHz) which is base 44.1KHz clocking.

Update: With USB input i mentioned above is correct. Vivaldi doesn’t have DSD/48 so everything played in DSD always locks to 44.1KHz.