Star Lost - Low HFD events

[Robert Lowenthal]

While guiding using an OAG with an ASI174MM Mini on a short-focal-length refractor (400 mm), I’ve noticed that PHD2 occasionally auto-selects a guide star with HFD around 2.0 px that later generates intermittent “Star Lost - Low HFD” events, even though other suitable guide stars with larger HFD values were available. My Minimum Star HFD is set to 1.5 pixels.

Would it make sense for the auto-selection algorithm to consider a candidate star’s margin above the Minimum Star HFD threshold as one factor in its ranking, rather than treating Minimum HFD only as a pass/fail criterion? For example, if two stars have similarly good centroid quality, would there be value in preferring the one with greater HFD margin to reduce the chance of transient Low-HFD events?

Also, is there any downside to intentionally defocusing the guide camera slightly to increase guide-star HFD from approximately 2.0 px to 2.5–3.0 px while maintaining a very high SNR (around 200)? My thinking is that a modest increase in HFD might provide additional robustness against Low-HFD events without materially affecting centroid accuracy.