A few weeks ago I changed the 45MHz modulation depth in full lock to make measurements similar to LLO 42731 and 48595.

The attached plot shows the POP 90 counts normalized to the nominal value.  (The userapps/isc/h1/scripts/step_45.py script that changes the modulation depth automaitcally scales the gain of POP 90, so here I divided out the gain before normalizing).  Doing the linear fit gives 85mW of light entering HAM6 that is not 45MHz, based on the AS_C calibration.  

In 48595  we found that there is a signiifcant amount of 9MHz power at the AS port, 24% of the non 45MHz power.  I didn't change the 9MHz modulation depth  this time, but if the ratio of carrier and 9MHz at the AS port has stayed the same, we would have 21mW of 9 MHz power at the AS port and 64mW of carrier.  We have 20mA on the DCPDs, with their responsivity of 0.858A/W this is 23mW of carrier power which is transmitted through the OMC. 

This gives us an estimate of 41mW at the AS port which is carrier and not mode matched to the OMC: 85mW - 21mW 9MHz - 23mW carrier transmitted by OMC. 

Although we don't have an estimate for the OMC mode matching, we can use 41mW of carrier HOM to estiamte the contrast defect at the BS similar to Valera's estimate, 41mW/(33.4W at prm *power recycling gain 43.8*Tsmr 0.3234) = 87 ppm