Maybe, all here hearing a difference, can post the way they operate Roon.
Core, networked, USB attached, what DAC. Maybe we get some common criteria.
My core runs on Linux Ubuntu 20.04 on a virtual server. 4 GB RAM, 2 CPUs.
No difference
I’m 100% sure there are no differences, it is a psychological thing (expectance bias). Happens all the time (and for every release). Same reason new cables which have zero relevans sound different. I’m also 100% sure you will not be able to tell the difference between Roon 1.7 and 1.8 if you are not aware of which one is playing 
Hi Olek,
that was exactly my hearings and feelings after hearing my well known favorites!
Great work roonies !
Bassman Ralph
I think people’s observation (including me) is that the background noise during signal reproduction has gone lower. It does not say anything about « background » noise present in the absence of digital signal, such as whatever residual humm or hiss or thermal noise from tube stages that one can hear close to high-efficiency drivers. I use also efficient drivers (from 100 to 105 dB/W/m).
RME ADI-2 DAC has built-in bit perfect test. It passes with Roon 1.8, same as it did with 1.7, meaning there is no technical reason to sound different. It is exactly the same, original signal sent from Roon Endpoint to DAC. Also it does not pass when I enable Parametric EQ or volume leveling, which is expected.
But nothing has changed.
I do hear differences whenever the air makes some big changes like tempreature and humidity though. Like I said before, and that wan’t a joke, it’s freezing outside and everything sound great overhere. but most audiophiles don’t even consider that air has probably anything to do with audio 
They do believe that they can hear the difference between a cpu computing a cycle too many or between -120dB and -124dB noisefloor 
There have been many psychological studies about expectation bias. I have read many of them over the years. Why? Because I wanted to learn not to fall into the trap of it. Why? Because I can’t do my job reliably when I let myself lead by expectations. That’s not as easy as you might think. Since then, it’s remarkable how little difference I hear between software players. Maybe I can’t hear it anymore because my expectation is that I don’t have any expectations, maybe I don’t “believe” it anymore. Oh boy, I’m trapped in a vicious circle.
I have Intel NUC with Roon Rock further Sotm SMS-200 further DAC for USB.
Nothing changed in the system during the upgrade.
Agree 100%
I think the USB connection is a common factor in people who report sound changes for better or worse.
Clearly a weak point as you are not isolated from the core. So, if the core is busier, it will generate more noise, hence issues.
With a network, Roon tested or Roon ready solution, I would think these issues do not apply.
Don’t you find it strange that almost all users who are unhappy with the new sound describe the shortcomings very similarly? Can psychosis be contagious?
A null test would easily put this all to bed.
Record the output of a DAC from roon 1.7 vs 1.8. Invert the phase of one of the recordings and play them both at exactly the same time. If there is a difference it will be audible. If there is silence then the 2 recordings are identical.
Is Sotm SMS-200, which acts as an End Point, is it not an insulator from the core?
I know this is meant to be funny, but if listeners could see the volume dial, how is that a “blind test”? 
I think modern life has decisively answered that in the affirmative.
Another psychological test. Did you read all the remarks with the same attention or did the remarks that confirmed your thoughts get a little more time and interest?
I read a lot of opposite remarks, from becoming thinner to fatter from dull to bright from less bass to too much bass. Not very consistent.
Have you ever noticed the behaviour of people who just bought a new piece of equipment. They ask on forums for reviews after they allready bought it and read only the positive ones. When in the end, in the case they don’t like the sound of their new equipment, they start reading again and notice the negative remarks in the reviews and ask themselves why they haven’t noticed them before.
I don’t know, but a network is…
Hi Nick.
What I can say is that with 1.7 I had to do a small digital correction of the bass (about 1.1 dB 40 Hz and below) to balance the overall spectrum. That is because the input sensitivity of my subwoofer is not exactly high enough - I use it at maximum volume but with an external RC crossover to increase its slope by 6 dB. The correction is computed as a wav filter on rePhase and input in HQplayer.
I was happy with +1.1 dB at 40 Hz and below on Roon/HQplayer with Roon 1.7. Having swapped to 1.8 and made the above comments, I have since then found that to find the same balanced spectrum, I had to add another +0.3 dB at 40 Hz and below.
Once that done, I no longer have that impression of slightly lighter and smoother sound that I had noticed. Overall the recording atmospheres etc… are as diverse as before in 1.7, but with a slight added clarity in low-level modulation. The sound is as involving as before, slightly cleaner : a combination of even lower noise floor and with further increased spatialisation. The silky texture of analog tape hiss of well mastered analog recordings is really well preserved. So, happy.
To reconcile these observations, I would be tempted to say, because of the A/B comparison to HQplayer alone is flat and remains flat, that for whatever reason, Roon 1.7 did introduce about 0.3 dB emphasis below 40 Hz, and was not exactly as clean as Roon 1.8 is, in the way it streams the input. The perception of that change might vary - people with slightly heavy bass might see the 1.8 as directly better. A pity I did not think of the HQplayer (alone or with Roon) test while I was still in 1.7.
Now I do share the opinion that it would be good to have some more clarifications by Roon Team about these topics, just because we are quite a few sticking to Roon (or Roon/HQ) primarily for the audible quality of its digital processing stream.
If 1.8 is really flat and 1.7 had a very slight bass emphasis, it would be fair for them to confirm this. We can’t easily go back to 1.7, they certainly maintain some comparison system.
By the way, on my system with a Roon Core on MB Pro laptop linked to DAC via a galvanically insulated 80 cm USB cable, all the library is accessed via Wifi on local network. If that may matter for comparisons and because that way of working limits the amount of high frequency digital noise in the analog system.
How do you imagine that? I don’t have version 1.7 anymore, unfortunately.
Above was the link where it said that the sound was “optimized”.
If this optimization doesn’t affect the sound, then why did they do it?
Noting that I have the underlying assumption that HQplayer itself is « flat » by design, with the PCM filters I use (sinc-M, no dither).
I am using a Antipodes S40/S60/EX combo with a Auralic Vega G2 and after the 1.8 I did hear less bass and dynamics impact. I did known regarding the 1.8 update, however, I didn’t know regarding the sound quality issue some people are having. I didn’t expect to hear worst, contrary, my expectations were to at least have better sound quality. This means that in my case, there wasn’t any psychological thing.
Due to my experience with the Vega which sound different after a FW update (I always detect that without knowing that was a update), I just turned off the S40 completely (I also disconnected the main PC but maybe not necessary) and the bass and dynamics came back. Unfurtunately, I loss the Ethernet (11:30 pm) and wasn’t able to keep trying other tracks, even from the local server, roon started to act very weird. I will do more test today.
Carlos