Roon 1.5 Sound Quality

Same setup for me except for the speakers. Nice!

V1.5 sounds great to me, especially the MQA decode/unfold. I like the main DSP Engine “on/off” switch in addition to the individual DSP switches. At times, I want to just quickly disable everything with one switch and not have to switch several different switches to accomplish this (if there wasn’t a single DSP engine switch).

I’m wondering if some of what folks may be hearing is related to perhaps extra processing that the new Roon version is doing and so this could somewhat effect the sound. Different DACs would be more or less susceptible to this. I’m using a Chord Hugo 2 with iFi Galvanic 3.0 and also a Devialet and an iFi Retro 50 and other noise reduction devices and my experiences match what Rob Watts (Chord Hugo 2 designer) has stated, he doesn’t really notice much change with various means of cleaning up the USB signal on the Chord (the iFi Retro and Devialet did benefit though). Noise (likely mostly conducted in) is another matter and is always relevant. Still a bit of a mystery to me what is actually affecting sound and what isn’t.

I use the Roon card in my Mytek Manhattan II but before the Ethernet cable enters the card I have the Acoustic Revive RLI-1 network filter! A must have device! No DSP in Roon Dac handles all the MQA. It’s definitely a tonal change in this upgrade. Not saying it’s bad but it’s tilted up ever so slightly!

1 Like

When I turn off all DSP and compare the same track played with Roon vs. played with BubbleUPnP, I can tell no difference.

It’s an easy way to toggle the DSP engine as a whole without losing the state of individual enablements.


As roon gets better at sound prosessing our music
The uptic will seem britier on some systems
But it also sounds better on better systems
As for output I still use hq player ona second desktop.

Same here!

It’s my impression, too.

6 posts were merged into an existing topic: Yanni or Laurel or MQA?

I follow my system changes very closely. 1.5 clearly changed sound from 1.4 to me, and in a bad way. Only a couple of days before a friend came round and I said he had to listen as my system was sounding seriously impressive on 1.4 (and had done for weeks and weeks). Then 1.5 upgraded a couple of days later, and the magic was gone, it’s a bit harsher, more digital sounding, less smooth and in a bad way. Before it was sort of perfect, very analogue sounding, now it’s more forward, digital and edgy. Oh well, hopefully the magic will return in a future update.

FYI I use Tidal mostly, upscaled to DSD256 using “7th order clans” and “precise, minimum phase”. Non-mqa material (obviously that’s most of it) is sounding worse in the way described. Mqa material is not worth commenting on as there’s different stuff going on there so it is impossible to compare, but it also carries that same signature of sounding a bit forward and harsh.

I hope Roon sorts everything as at the moment it’s sounding slightly “cheap” in comparison to how it was sounding.

Nothing else had changed in my system in months, so the change was certainly down to Roon 1.5. I wish I could downgrade to 1.4 as I would in a heartbeat.


Agreed! Not as musically enjoyable. Become a little to “digital”!

1 Like

If you use TImeCapsule on Mac machines, easy to downgrade or upgrade as the previous version is accessible in backups and can be reinstalled in a few clicks, and compared back and forth, with simple renaming to avoid confusion.
I actually just did that to check for a doubt on what exactly was different between HQplayer 3.20.4 and 3.20.1, following an upgrade. Without it, like for Roon, the site does not allow for easy access to previous versions, which is a pity.

Then, like several commenters here, I find Roon 1.5 an audible improvement over 1.4. And this improvement gets further opened with HQplayer. In my case I have lots of precaution taken from electrical to mechanical and in that setting I am happy to use poly-sinc-mp associated with Shaped dither, as a more recording-robust alternative to no dither at all. I am probably in the purposely favorable case of careful and somehow rigorous assembly.

As soon as the system gets sufficiently resolving to the kind of subtleties involved by incremental software improvements, a number of precautions must be taken, so that as lower levels of signal can be unraveled, it does not bring in additional noise, harshness, be it in analog or more insidious digital form of induced jitter for instance.
Imagine you change the assembly of a precision balance, with a new almost frictionless articulation. Will you make more accurate weighting ? Not necessarily, if for instance the chassis of your balance lays on slightly unstable support, you will start to have non-repeatability that was not observed previously.
The exact same happens for audio, and the analog end of the digital audio path is quite sensitive to HF noise as well as, say, a ground reference that carries some residual potential, which can induce jitter. Jitter risk increases as higher sampling frequencies are being tackled, by the way. The choice of dither is as important as the choice of filter, and a progress in software, hardware or even electrical environment can mean that another optimal setting can be used.
So hopefully a better software will help you find your way towards a better fine tuning / configuration / element selection. This requires some patience and rigor - and quite some time, also, so as you said it would be good to be able to downgrade easily, let alone for verification, as I just did for HQplayer.

I totally agree, to the point where I can’t listen to my main system anymore.
I just don’t see how it’s possible and I have gone through my system and checked all connections. I just can’t believe it’s the software that has made the difference but I am definitely not enjoying listening to music on my main set up since 1.5.

MQA sounds too bright. Just compare any Tom Petty non MQA to the MQA version. I don’t remember it sounding quite as bad with the unfold in the Tidal app. Could the MQA update be causing folks to hear some of their collection now unfolded to 96KHz instead of at 48KHz and commensurately brighter due to the hocus pocus Disneyland rubbish minimum phase filters in MQA?

Minimum phase filters have no place in audio playback as far as I am concerned…messing with relative relationships of various frequencies just ruins natural timbre. Why has our industry forgotten basic common sense audio engineering principles - Linear Phase was and still remains the golden rule.

Tidal is really slow lately.

I use local files only for serious audio listening. In my experience, even a bad local network can generate harsher digital music reproduction than a clean one - jitter is not a pass or stop issue !

I would be very confused in a situation where simultaneously:

  • I have slow internet access to online music
  • I have made a software change
  • I find the music getting harsher, less enjoyable.

Because timing is essential in music and even asynchronous data such as online music has to be tailored in the end, to the restitution via the DAC, using some buffering usually.

Yet the best way to warrant the smooth buffered output is to enslave the delivery of digital stream to the DAC itself. And it is way easier to do this in a local wifi network - or directly in a machine - than in a setting where the server is thousands of km always and goes through large pipes, in competition. with millions of other data streams.

I would therefore recommend to test all these updates quietly, only on local files, to remove one potential cause of digital harshness and be reasonably sure to maire repeatable tests (AC quality can still fluctuate). Also, why not do a complete system restart as well and avoid starting many other software tools that will compete for RAM usage and bandwidth occupation, therefore potentially stalling the music flow and inducing jitter ?

PS I have no experience in WIndows-based systems that I have long stopped using for audio purposes, going the safer, albeit more expensive, Apple route for digital audio.

A post was merged into an existing topic: Linear Phase v Minimum Phase?

Same brightness on local hi-res and non MQA Tidal.

Then you problem is elsewhere. But that was worth checking.
Sorry, no experience of MQA to date, although I read articles and understand the concept. I have enough to play with between 44/16 AIFF and various High Res formats.

I guess you ompare identical formats only - not CD Tidal versus local HR for instance. I would ask the Roon guys then…

It’s bright using Roon not elsewhere!