Yesterday there was one attempt at doing teh CARM offset reduction, ndscope shown. At these early stages of lock acquisition, we should not see dips in the reflected power, I think.  When we were in the CARM_150PM state, we had 4 times the single arm build up, but the reflected power had already dipped to 95% of the off resonance power. You can compare this to the plots in 43344 and see that the recycling gain must be low compared to what we've used to lock in the past.  We can attempt to adjust alignment to increase the refl power while we have the arm locked on TR_CARM.