I measured the common mode rejection ratio of the interferometer to intensity noise.
CMRR is roughly 1/17 @ 100 Hz.

Method:
 1. Measured the transfer function from K1:PSL-ISS_FIRST_SERVO_PDA_INF_OUT_DQ to K1:OMC-TRANS_DC_SUM_OUT_DQ while ISS is not on (the same timing as Yokozawa-san's measurement in klog #12983; 15-Feb-2020 at 18:30 UTC).

 2. Dividing the transfer function by 8 (8 mW offset in TRANS_DC_SUM) will give you the transfer function from RIN of IFO injected beam to RIN of OMC DC PD. (Actually, K1:PSL-ISS_FIRST_SERVO_PDA_RIN_OUT_DQ is the RIN channel, but INF channel is almost the same since normalization is done by 0.995 or so, and RIN channel was not DQ-ed at this time).

Result:
 Attached. Measurement below ~100 Hz is not very reliable due to low coherence (shaded region shows the error).

Discussion:
 Although there's a finesse asymmetry between two arms, more common mode rejection is expected at high frequencies considering CARM cavity pole.
 Flat RIN coupling at high frequencies suggests that intensity noise coupling is mostly from MICH contrast defect.

Next:
- Measure CMRR for frequency noise (CARM coupling is already measured (klog #12998), so what we need is a calibration of CARM error signal to Hz).
- Compare them with the simulation in JGW-T1910352