I just received a UR22mkII about a month ago, and I have exactly the same issue using v1.9.6 and v1.02 of the firmware. I am running Windows 10.
It has been tearing my hair out and I had literally tried everything I could think of in order to fix before stumbling across this thread (thanks for all posts so far). I was just on the brink of raising a ticket with Steinberg. The laptop is a fresh install and has had all the bloat from Windows 10 stripped away (as much as possible anyway!). Being my first ever interface I initially thought it was ASIO buffer size related, or some weird performance spike causing the issue. But I couldn’t see how because the laptop has next to nothing installed on it and is highly specced (Core i7, 8gb RAM, 128gb SSD).
Given that v1.9.4 of the driver is not compatible with Windows 10 and that people are upgrading to or installing Windows 10 more frequently now, it would be useful if Steinberg can acknowledge this issue and work on a fix. The drivers with the issue (1.9.6) have been around 2 months already so i’m surprised they haven’t been pulled. I don’t want to have to return the interface if this is not a hardware issue, because apart from the driver problem it works as intended and the latency is good enough for practicing my V-Drums - I get no pops or crackles on the smallest buffer size of 64 so I can’t ask for more really.
Every time I practice on my e-kit, at some random point this distortion is guaranteed to creep in at any stage so it is very easy to replicate if more samples are required. As sg2015 says, this is regardless of buffer size, it is present on both outputs and lasts a random duration. It also happens whether loopback is switched on or off. Changing any ASIO settings to provoke a reset or stopping playing for several moments fixes the problem in the short term, but the issue always returns eventually.
Hope Steinberg can get this resolved ASAP.