Frustrated with Roon's sound quality vs Lunin Minimserver app

I just installed Roon and the streamer/DAC Teac NT-505 in my audio system.
Here is my system configuration: Roon core runs on Dell I7 with music stored on SSD, 35ft Cat5e ethernet connection from Comcast Router from my office upstairs to my HT’s Asus router, then another 35ft of Cat7 cable to the Teac 505. The Teac then feeds Conrad Johnson preamp and Wolcott mono amps.
I also run Minimserver from the Dell to stream music to the Teac using the SAME connection as above.
I performed the following A/B/C tests:

A. Dell NAS/Roon/Ethernet/Teac: The sound is warm and somewhat veiled/ muddy, no clear separation of instrument.
B. Lumin app/Teac/USB flash drive: Best SQ.
C. Dell NAS/Minimserver/Ethernet/Teac: Better SQ than option A using Roon.

How can I improve SQ of Roon? I just subscribed Roon and want to keep it, but if SQ is not as good as other options, I may have to cancel it.

I have about 1000 albums stored on NAS, and don’t currently use any streaming services.
Thanks for any input.

Turn the volume up 3db when listening to Roon.


When I was trialing Roon, I used a $30 raspberry pi endpoint with usb connection. At the time, to my ears, it sounded better than my Aurender, also usb connected, I was amazed, and hooked! Occasionally people post X sounds better, but there’s never really any conclusion why, so I’m interested to see the responses here.

It is not necessary that Roon has better sound quality, but that has been repeated so often that it is almost gospel. Have a different setup but found that Minimserver was better than Roon via Airplay or optical to a pair of Devialet Phantom Golds. Decided not to subscribe to Roon after trial…

Roon encourages this thinking though, with their recommendation that the core be separated from the output and I’m sure I’ve even seen recommendations that closing the remote window when not in use could also affect sound quality. Both are untrue, in my opinion.

1 Like

Try swapping your last cat7 cable for cat5e or 6. cat7 are grounded lan cables, and they maight be injecting noise into your system, connecting your router ground to your TEAC ground (that is one possible explanation o perceived better SQ on a USB stick than any of your solutions via ethernet).

Also, as said before, be sure to exatly level match your minimserver and roon. You may use audiotools on android. nice app, and works well.

Finally, should you have one at hand, or the possibitliy to borrow from someone, try a different renderer. It may as simple as a respberry (for comparing roon to minimserver), more elaborate as a SOtM or a Sonore Rendu ,or the fanciest type like a dcs bridge or an innuos server. Should you get a rendu, you could try the Roon+HQplayer solution.

My experience is that, with the psaudio bridge (in the directstream dac) upnp (minimserver) was about the same as roon. Moving to the rendu line, HQPlayer was way better than roon or minimserver. Then, moving to the dcs vivaldi, all things are equal again. My guess is that the “RAAT” protocol implementation makes all the difference. Never tried the SOtM renderer.

1 Like

I wonder if it would help to get a long 80 foot cable to connect directly from Comcast router upstairs to the Teac downstairs, this is a temp solution in trying to isolate the issue, I can not run a cable that long through my living space as a permanent solution.
How do I ground both the Asus router in HT room and Teac?
I don’t have any other streamer unless I buy another one. I don’t think the issue is with the Teac streamer since both Minimserver and Roon go to it, but the sound from Minimserver is more transparent.

cat7 ethernet cable is not isolated. That’s the point. The metal plug works as ground, tying the router chassis ground to the ethernet jack (chassis ground) of the TEAC. In my view, you do not neet a 80ft ethernet cable. Just swap the last 35ft run of cat7 for a common cat5e (it should be pretty cheap) to isolate this as a factor.

For you to avoid such grounding problems, your router would have to have a ground lug. That is not common in consumer routers.

1 Like

there are a lot of discussions about cat7 cables. look here, for example.

But I agree ethernet in cat5e or 6 is galvanically isolated. So you should not need to replace both cat7 runs.

1 Like

I don’t quite understand how Roon SQ would improve if I replace the Cat 7 cable with the Cat 5e cable while the SQ of Minimserver was better even with the Cat7 cable.

You are 100% right.

I am proposing to you 2 proof-of-concept tests: the first one is to know why USB sounded better (for it should not) - for this, the test would be to swap the ethernet cable.

The other one is minim x roon. here, I would ensure the level matching (via a phone app). Then, even if level matched you continue to prefer minimserver (which, by the way, is great), you could: a) continue to insist with other endpoints, with better roon implementation; or b) see what matters to you most - roon user experience and convenience, or minimserver SQ.

I went by the minimserver route until I came accross the vivaldi. then it became a non brainer for roon - same SQ with better user interface.

Anyway, should your fist test be well suceed, at minimum you would have a better ethernet setup, be it minimserver or roon.

Hope it helps.

I would recommend you post an image of your signal path when using Roon and device and audio settings. So we can see if anything is obviously a miss in your set up.

To show signal path click on now playing area and the little dot to the right.

1 Like

This reply from Andrew from dCS might help clearing things out a bit.

I am new to PC audio/streamer. Please educate me how 2 different source signals feeding one end point via Ethernet result in different SQ could be the issue with the end point. I would assume it is the issue with Roon vs Minimserver audio signal upstream.

It’s called an ‘Auditory Hallucination’.

Or it’s the way the code is implemented, as I understand that your upstream chain is the same to minimserver and to roon.

You see, the code to act as a roon endpoint is different from the code to act as an upnp endpoint, and also is different from the code to act as a HQPlayer endpoint.

For equal level implementations, one should not hear any difference in SQ. That is basically what Andrew is telling us, from a dcs perspective (that matches my opinion). But I have no experience with your endpoint/streamer. Apart from different loudness (a louder source will be perceived as better), I can only imagine that might be some difference in implementation between the roon code and the upnp code, that favors the upnp.

I am not very knowledgeable also, and I can only share my personal experience and the experience of others that I see in different fora. I can tell you that, for my ears, no difference in PSAudio (roon x minimserver), no difference in dcs (roon x minimserver), hqplayer is better in sonore ultrarendu (roon x minimserver x hqplayer).

I replaced Cat7 cable with Cat5E cable, no SQ improvement with Roon, even worse both Roon and Minimserver streaming were not stable, sometimes dropped.
I am not sure what will be my next steps, I may get a Bluesound Node 2i and use it as a streamer to feed the Teac DAC.
Or maybe just cancel Roon subscription and use Minimserver to play music. I value SQ higher than better UI.

You need a switch between your NT-505 and your office. Close to the NT-505, maybe with a linear PSU. If it were me I’d pull it all in to the same space with short hops and all powered with LPSU’s. That is how I got the best sound out of my setup together with a specialist network card on my server (Jcat Net Femto). I can’t explain why this might work, but your best sound is playing everything locally and the bigger you make your network the worse it sounds.
My setup is internet router >>Switch. The switch is connected to both my server with the JCAT card and my NT-505 by short cat8 patch cables (they are thin and neat). All music is held internally on my ROCK server. i7 8700 processor, 8gb RAM, completely silent.

Finally, what I will say is that my system came together over a period of time and I kept my server away for a lot of that time. But after I got the 505 I got decent gains from pulling it all together. So apart from the Jcat card it was a matter of reconfiguring what I already had. I wouldn’t expect you to go out and buy a new server to try my theory, and if I were you I’d probably cut my losses and go with minimserver and the Lumin app.

Thanks Henry for your assistance and input.
I don’t get what is the Switch. I have internet router with a long Cat5e cable connected to another router in my HT/audio room so what will the role of the switch? I won’t move the Dell server to my HT room, it would be very noisy.
Have you tried HqPlayer? I just got it installed, will audition it with Roon tomorrow.

The switch would serve to isolate the streamer from the effects of a relatively long run of cable. If the problem is distance, it should help. Some companies take the idea of the role of a switch further by producing audiophile switches but I haven’t tried them and wouldn’t necessarily advocate doing so untried. The problem is of course the options to try before you buy are limited in many parts of the world. I accept that much of this goes against a lot of the accepted wisdoms but I can only speak from my experience.