Sound quality of Roon


I am considering trying Roon.  I have been using my Bluesound Node but I am going to upgrade as I do enjoy streaming more and more using Tidal.  It is quite an investment to get a NUC or Nucleus and then have a separate tablet to control it all.
 

But apart from the cost I have read some people say Roon does not sound good.  Their streamer by blah blah sounds better.  Is this true?  For all that is required to use Roon, the hardware, the subscription and all, would Roon be popular if it made digital streaming sound bad?


I would love to hear people who have experience comment on this.  There is info on the Roon Labs discussion site but as you can imagine it is saying this is BS Roon sounds great.  I guess Roon as a software also has had updates, so maybe this is a thing that might have been true in the past?  

troidelover1499

I appreciate the response and pointing out the two items you mention @blisshifi  and those are the two everyone says, noise and timing, but are you saying that you can hear music slower/faster or with more or less noise based on streamer? Who else can hear that?

@sonicfanatic "Timing" does not result in music going faster or slower - it is more about the interpretation of the data which as a result makes imaging more solid, or in the case of poor timing/clocking, makes the imaging sound more flabby or loose. So improved timing simply helps the signal produce sound that snaps/gels together.

The same paradigm applies to photography. Think of it as a lens that can produce an image that is completely in focus vs one that is just ever-so-slightly off focus that you might not really notice unless you had the two images side by side. Or even two lenses that may be equal in quality (Canon vs Nikon top of the lines). When people shoot with them, they perceive those images to be the best reproduction to what their eyes can see. But sometimes those manufacturers can bring new innovations to those lenses that improve upon the focus and clarity, even if used on the same camera body which produces the same image resolution.

Timing is one issue, based on clocking from the source (server/streamer), in the DAC, and the cable in between. For more fun, different digital interfaces (USB vs SPDIF) prioritize their clocks from different places. USB uses the DACs clock, whereas AES prioritizes the source’s clock. Some DACs will still reclock a SPDIF signal, and this can result in both great and poor outcomes, depending on how good the clock from the source already is. Many people don’t like the USB interface because it tends to be noisier, even if the DAC offers incredible clocking. This is typically due to lack of isolation in the ports and circuitry, and the fact that most USB cables are powered, carrying 5VDC in parallel conductors with the digital signal. So what one might gain from clocking they may lose in noise.

Noise is a whole other issue. Much noise in digital is indescribable. It’s not that you hear the noise like white noise, but it does cause some distortion to the output and you will know when you hear less noise in comparison. Lower noise reduces fatigue when listening over time, and with digital sources lower noise also tends to shift the tonal balance to feel more natural because it tames some of the fatiguing peaks and allows you to hear more, easier, and do so in a more relaxed way.

With this in mind, even though SPDIF sources are generally limited to 192khz resolution, many people prefer this interface because of the improved clocking and reduction and noise compared to a higher resolution signal such as a USB which can deliver 768khz of resolution, but at times with shortcomings due to noise and timing. Some DACs are designed without USB in mind because of this principle, like Berkeley’s reference which are around ~$25K, or the Aqua Formula xHD at around ~$13-15K. They prefer to offer a max resolution of ~192khz.

Thanks again @blisshifi  for the explanation and the two factors you raise are imaging of the sound towards more life like and fatiguing peaks which I assume tends towards the higher end of the frequency response. Are others hearing these differences as a result of a streamer change and not related to dac or speakers? Or have you had the experience where the streamer does not not appear to make a difference? How expensive did your system need to be to hear this difference? Did you have lower cost systems before where you could not hear the difference?

@jji666 wrote: 

I definitely agree that trouble-shooting Roon is a highly complex process. But it’s just not fair to say that nearly all problems with Roon are the user’s fault and usually due to their crappy network. True: many users that complain of problems have crappy networks (many are powerline! Yuck!). False: Roon is stable if your network is.

There are many ways in the way of ordinary use of Roon that cause it to tax itself to the point it piddles. Just do 15-20 searches in a row, adding tags to your search findings in between, then bring up a large tag and try to use focus on some of the tag results. It’s gonna choke.

Someone mentioned that having many open remotes also seemed to tax Roon. I’ve had my eye on that as a potential cause as well and there does seem to be some correlation.

I’m definitely not kicking Roon from a software design standpoint. I couldn’t do what I do with my music room without Roon. It’s brilliant in many respects. But that doesn’t make it stable.

AND, if it were the case that a restart of the core software temporarily refreshes IPs and other data that were causing problems, why doesn’t Roon just do that every so often?

I agree mostly.  I didn't mean to imply that Roon is 100% reliable if you know what you're doing.  Every situation is different.  But I also disagree that users without issues are just 'lucky'.  Your point on several open remotes as a potential cause could have merit.  I use a single tablet and always close Roon out whenever I turn off my DAC.  I don't even do it for Roon purposes.  It eats up my Samsung's battery when left open/running!

A restart should definitely not be an automated process.  Not everyone has IP conflicts.  In fact, I would bet the vast majority do not.

It seems to me reading all of this that the computational power required for full Roon implementation is per se detrimental to good sound quality. Hence the need for separating the (noisy) core and the clean end point. Other than the associated cost and cable dependency the question arises why going with the slimmed down processing power of say an InnuOS server wouldn’t be the more elegant solution from a sound quality perspective, fully acknowledging Roon’s superior customer interface.