Keita K. informed me that whitening for OLs were off for both ITMX and ETMX and after looking at the spectra, I also agreed. So I went out to the field racks to check that the settings were correct. The OpLev whitening boards use a binary I/O daughter board to enable the gain stages. They looked like they were on the correct settings for two levels of whitening, which matched the de-whitening filters set in software so I unplugged and plugged the board back in and this seemed to solve the problem. I'm happy that this solved the problem, but I'm perplexed on how it happened. The daughter boards just plug into the front of the chassis with no screws so it may have been bumped loose, not good. I'm still investigating when this might have happened but it has some pretty big implications for the relatively new damping scripts that Stefan made, I've sent him an email.
The hardware whitening, which is supposed to be 2 stages of 1:10, was off while the software awhiteners were on for ITM. After Thomas plugged the daughter board off and on, both of the whitening came back.
This was first found while measuring the WFS sensing matrix, as there was a 90 degrees phase difference between WFS and oplev when we were wiggling ITM at 1Hz.
This means that ITM OPLEV was underestimating the angle for ITMX. This is nothing at 0.1Hz, a factor of 1.25 at 0.5Hz, and a factor of 2 at 1Hz.
I don't know if the situation was the same for ETMX, though I know that the ETMX OPLEV whitening changed after Thomas did his trick of plugging out and in. I just don't know if it was no whitening or just one stage whitening or what. That is to be confirmed by Thomas later.