This is to allow us to have a small amount of power going into the fiber coupler without having to reject so much power with the half waveplate and the beam splitter cube. The mirror labeled R=10% but it actually reflects 5% (AR coated for 1064). The coupling efficiency through the fiber coupler is not great at the moment (~20%) but that's plenty enough for us. I set the power back to where we normally operated (~0.02 mW reflected, 0.5 uW transmitted). I think we are ready to have another go at low CLF power.