Upsampler vs dac filter settings

My system evolution unfortunately stopped a while back, but still fills us with joy most evenings! It’s a mix - Scarlatti dac, Vivaldi Upsampler, Rossini clock. Its clocking works in a compromised fashion as I accept an upsampling regime that is fixed - in my case to 352.8. I use Roon and stream a variety of file types that might be based on either 44.1 or 48 bases. Not ideal I know.

Currently I’m using filter 1 on both dac and upsampler, but have failed to understand how a filter selected at upsampler level might affect the dac. As far as I can deduce it appears 5 filter are available for 44.1 > 352.8 conversions but just 1 for 48 conversions.

Advice and recommendations greatly appreciated

After upsampling it is necessary to use a filter to correctly reconstruct the file before it is sent to the DAC for conversion. Filter choices are, if reasonable, a matter of subjective preference. The extent that the result is audible will therefore differ from circumstance to circumstance and person to person. Frankly I cannot discern much if any change moving from filter 1. Others may find it a major issue for all I know. Filter choices at DAC level are another matter for me and very audible.

Yes, your reading of the number of filters available at different sampling rate is correct ( see p.29 of the user manual).

My quibble is with your use of fixed upsampling rate of 352.8 for both original sources of 44.1 and 48 sample rates. |The upsampled rate should be an exact multiple of the sample rate i.e.for the highest rates, 44.1 fKS/s files should be upsampled to 352.8 but 48 KS/s files to 384. However I think you are aware of this.

1 Like

Peter, thank you for your input. On the filters I can’t understand for example how selecting a filter on the upsampler that has a sharp cutoff might then work with a filter on the dac with a much shallower cutoff: to me the job has been done by the upsampler.

I note your comment on my upsampling 48 to 352, as far as I can see there is no way round this (automatically) as my Rossini clock 3 out goes to the Scarlatti and is fixed at 44.1. The manual does note that Aurender supply a dongle for their A20 that effects the automatic change, but I’m not sure if dCS has a similar solution for a part legacy system such as mine.

I think that you are right though you can change the clock frequency manually to 48. It is too long a time ago when I used a Paganini ( same as Scarlatti) to remember fully but it had automatic frequency changes on USB - the port being on the Paganini (or Scarlatti) clock however. I guess that this mix of dCS generations has a number of drawbacks.

Yes, I had a similar functioning system of u-clock with Paganini upsampler and Elgar dac. It automatically switched. Shame the Rossini doesn’t.

Beginning to think it’s time to reconsider my compromised system given that any dCS move has massive financial impact.

I am going to change my mind now from listening to to my Vivaldi Apex. I could hear little if any difference originally but the Apex upgrade is very revealing and I am now ware of changes when moving upsampling filters. Not a night and day difference but it is able to produce small but real preferences.

For me it is only one small thing demonstrating the superiority of Apex which, now burned in, continues to amaze me. So far I have not played a single familiar piece of music without new aspects being revealed.


Indeed Pete, that goes for my now burned in Rossini Apex DAC too!