I think I hear a difference between projects played in Cubase 10 and 10.5. 10.5 sounds clearer to my ears. It’s hard to believe, and I guess I could be fooling myself, but I wonder if anybody else has the same impression.
Matthias / Fabio - have there been any changes to the audio processing in 10.5?
R
No. A thousand times no. 
(Unless you have some setting that’s causing it. )
There could well be something that’s causing it that’s not the audio engine, but I’ve no clue what that is at the moment!
Richard you are right 10.5 sounds better …
After years of development Steinberg engineers come up with new kind of binary markers called FAT Zeros
FAT O make digital music recorded on Cubase 10.5 sound fat and richer
Most likely you fool your self. However it’s easy to test and do a “null test”. Do the same thing in both. Take the output and invert one and see if they cancel out. (You can not use plugin with random behaviour like things with noise and start points of LFO’s etc)
Unfortunately all my tracks use soft synths and I suspect they won’t null because of the random elements - but I may give it a go!
Well, I understand digital audio quite well, so it does seem strange. There is probably some difference in settings somewhere that I have probably overlooked.
You can make it work buy freeze the tracks that might has some randomness before mixdown and use the same freeze on both.
Ok, good thought!
There are areas that are not well defined in VST2. For example automation points are not sample accurate as they are for VST3. Are you using automation where you think there is a difference? VST2?
No…
This has been reported with every new version of Cubase since, well, since I can remember.
The simple fact is, there is no difference.
You paid money for something you know you didn’t really need, so your mind plays tricks on you in order to justify the outlay.
But I do like the concept of “fat zeroes” … wait till you hear “vintage ones”. 
Yes, unless there was a subtle bug that they spotted and fixed in this version that I was unlucky enough to suffer from. Unlikely, I know, probably the effect is psychological.
Fat zeros, my axx. It’s skinny 1s that make things sound pristine. 
I will see your “skinny 1s” and raise you 5 “quantum maybes”.
Different pan law?
Maybe bounce all those vsti to audio tracks and then use those audio tracks in 10 and 10.5 to do the test
Perhaps you changed in CB10.5 to 32 or 64 bit precision?
0’s, 1’s and O’s eh? Great work Steinberg! 
A slight change in volume can make a massive to the perceived sound quality - perhaps you had Control Room setup and it’s a different level or something?
I searched and found this because I thought the same. Doing my first session on it now. Seems smoother and more glued together.