The modern newton style telescope uses off-axis secondary mirror which could make the unfocused spot (a tiny ellipse) show the minimum HFR.
So the HFR could not represent the correct focus spot.
In my F4 Mak-Newt, the minimum HFR spot is always a vertical ellipse. I need to move the focuser a little out to get a proper round disk. This correction will only increase 0.1-0.2 HFR.
So I suggest to add roundness as for the Y value of the curve so that we can change monitoring from HFR to roundness when we like.
So the focus progress would be like that:
move focuser in one direction based on the HFR when HFR > 1.5 (a parameter)
received first <1.5 HFR
change Y value to roundness, calculate roundness instead of HFR
move focuser in the same direction and record roundness
received maximum roundness
move focuser back to meet the maximum roundness
I hope such function to be added into the focus module as an option.