This is a sister alog to LLO:72613. After discussion with the calibration group, Joe and I have proposed and pushed a small but important fix to the CAL_LINE_MONITOR_MASTER/COHERENCE block. When the IFOs are down, the coherence for various calibration and monitoring line transfer functions is 0 (this is good). However, the TF uncertainty, which is also produced by the coherence block, was found to be railing at 0 too. How can the uncertainty be 0 if there is no coherence?! It turns out that a coherence of 0 was causing a divide by 0 that, on the front end, presents as ... you guessed it, 0. The new h1calcs models (updated at LLO then synchronized by me to LHO) fix that. 

Now the coherence signal entering the uncertainty-calculation logic will be kept within 1e-6 to 2. We allowed for a coh value above 2 in order to avoid inadvertently 'hiding' egregious errors in the coherence calculation by forcing it to stay below 1. Thinking about this further now that I'm writing this...that wasn't necessary as the coherence channel itself isn't affected by our new changes. oh well.

I've attached an annotated screenshot of Joe's screenshot in LLO:72613 to show the new component we've installed. 

The h1calcs updates should be put in place during tomorrow's maintenance period.