Agreed about the 90 setting. My start was at 10 when the problem was observed. I tried the 90 setting to compare.Even at 90 out of 100 you are still putting 2K in series with your input, assuming the pot is linear - and the rated input impedance of the Fosi may only apply with power on - depending on what the input circuit is.
I think you just have to take the "it is what it is" approach - unless you are prepared to work out what the input circuit looks like - or ask Fosi to provide it for you.
The pots are standard 10k Bourns pots. The 2i2 output sees 20k with the tap to the V3. Granted the somewhat low V3 input impedance is a factor such that the effective impedance seen by the 2i2 varies more than would be desired, but at the 10 setting, even if the V3 impedance was zero, the 2i2 would still see 18k. So we have 5k parallel to what point the attenuator tap happens to be. At attenuator 10 the 2i2 sees 18k with a (remaining) 2k/5k parallel in series with the 18k. If I've figured this correctly.
Edit: I never measured the 2i2/V3 input when I had the 2i2 directly connected and got good results. I added the attenuator to the output to be able to set the 2i2 output near maximum and then attenuated to the desired V3 input voltage in order to run the 2i2 output with lowest possible relative noise level together with the REW added distortion option. That worked exactly as desired for the 2i2 loopback, but the V3 issue is a no-go unless I can solve this problem. It may not be solvable, I'll grant that.
Last edited: