Would a dedicated streamer help a Nucleus?

I’m running a Roon through a multi-room system but one audio setup is separated. This has Heresy IV speakers (and a KEF mini sub), Schiit Freya+ preamp, Denafrips ares DAC connected directly to the Core via usb.

I’m not getting a bunch of noise so I’m wondering if there would be any reason (noticeable improvement) by putting a streamer in this mix, like an Ultrarendu or Lumia.

Also just wondering what may be the weakest links in the current setup to the more learned minds.

Probably not. And you would be adding an extra point of failure. IMHO, I like to keep the audio chain as short as possible.


What a coincidence. I used to have a ultrarendu hooked to my denafrips ares dac. The nucleus was located at another room. I found out that hooking the dac directly to the nucleus, like you already have sounds much better.

I sold the ultrarendu on the second hand market

1 Like

This is very interesting.
Why would Roon recommend to split the core and streamer if it sounds better without?
I followed the guidelines and installed ROCK on two NUC:s, one core and one bridge.
Perhaps I should move the core from the basement and run it directly to my DAC?

Firstly, whilst some may perceive a audio quality improvement with a direct Core to DAC connection, it is by no means universal. Others may perceive the opposite and still more may perceive no difference.

I can think of two possible reasons why a separate streamer may be recommended:

  1. NUCs, in their standard form, have a fan present so, when worked hard (as in the heavy dsp case described below), may not be silent. Using a separate streamer allows the NUC to be located away from the listening station so that any fan noise will not matter.

  2. When running a USB DAC connected directly to your ROCK (Roon Server) you are effectively running two Roon software components - server and bridge with a local TCP/IP connection between them whereas when you use a separate RAAT based streamer, you are not running the bridge component.
    Thus, when you use a separate streamer, you are preserving more of the resources (processor and memory) of the NUC for the processing (including DSP) and distribution of music data streams to the various endpoints.
    This might not matter when you are running just one or two endpoints with low or moderate bit rate files and no DSP but could make a significant difference with multiple endpoints running heavy DSP and/or extremely high bit rates such that the Roon reported ‘Processing Speed’ approaches 1 for each endpoint.

Neither of these relate directly to audio quality although both could impact upon the listening experience in extreme cases.

No matter what your opinions on ‘the audibility of the quality of the USB data stream’, I doubt that you would notice much difference because of the architectural similarity of the two NUC devices - unless you are using heavy DSP and the NUC fan starts running or the extra workload on the now single NUC causes the NUC to fail to keep up with DSP workload causing clicks, dropouts and even stream failures as packets get dropped or unacceptably delayed.

At present, no matter how hard the NUC acting as the core is running its fan, you won’t hear it and the NUC acting as the streamer will only be loaded very lightly so:

  • it is unlikely that its fan would run at all (if kept free of dust).
  • it is unlikely that it could be blamed for any failure to deliver USB packets within the required time frame.

Thus I can see no mechanism by which moving the core and running the DAC directly from it could improve the listening experience.

Having said all of this, if you do not run multiple endpoints simultaneously and do not do heavy DSP, there may be other reasons (not audio quality related) for eliminating the second NUC. These, may, for example, include:

  1. Freeing up a NUC for other purposes
  2. Simplifying the network and thus making it easier to maintain
  3. Reducing power consumption