While passing through the Faraday isolator of the SDB1, part of the light is going towards B1p photodiode while the other part is reflected by the FI polarizer towards the ESQB.
This second beam is then reaching the polarizer of the last faraday isolator of the squeezing bench. In reflection of this polarizer there are a NF and a FF camera.

While looking at the data, we noticed that the minimum power on the cameras does not correspond to a maximum on the photodiode (see figure 1). This suggests that the angle between the polarizers is mistuned.
On the figure, the picomotor of the HWP has been moved of 6000 steps, so there is a mistuning of about 3000 steps. The steps are not calibrated but from our experience it is a relatively small angle.

I did some simulations (see figure 2) using the data points taken by Fiodor on the power reaching the camera (entry 43609), knowing that the input power was between 750 mW and 1W, and making varying the following parameters:
- The part of S-polarized light in the input beam
- The part of P-polarized light reflected by the polarizer of SDB1
- The part of P-polarized light reflected by the polarizer of ESQB
- The mistuning angle between the two polarizers


There are more varying parameters than data points. Thus many different possibilities.
However by keeping reasonable values for each parameter and succesively fixing a parameter and making the others varying, the simulation gives a maximum mistuning angle of the order of 3 °. 
This corresponds to a maximum losses of 0.3 %. This is relatively low, but it has to be taken into account while considering the total losses of the squeezed beam.

Retuning the polarizer of the ESQB with respect to the one of SDB1 would require to retune the other five polarizers of the other FI. This is a quite heavy task with risks of contamination of the optics.
It is not a priority task for the moment but it may have to be done later.