Since we have been struggling to understand our measurements of the IMC cavity pole, I had a look at the power in different ports of the IMC.  

The IMC visibility is 96%.  

MC2 transmitts 1.3mW, where 1.4mW are expected

Paul calibrated the MC2 trans QPD in uW ( alog 9716), there is a 90% splitter between the QPD and MC2, so with the QPD sum at 127.7uW-1.4uW dark offset we have MC2trans=1.26mW.  With 1.622Watts input power (measured by PSL periscope PD) we expect 1.622Watts *166( expected IMC build up)*5.1e-6 (MC2 transmission)=1.4mW

The cavity pole measurement (alog 13381 ) would indicate that the IMC build up should be 178, (instead of the expected 168)  which would require the transmission of MC2 to have decreased to 4.2ppm. 

The IMC throughput is also OK, based on Kiwamu's IM4 trans measurement yesterday (alog 13495 )and assuming the input Faraday has 3% losses, we have an IMC throughput of 90%.  If the cavity pole measurement were correct this would require MC3 transmission to have dropped from 0.58% to 0.48%.  

The simplest explanation here seems to be that the cavity pole measurement is wrong, and that the IMC hasn't changed.