Automatic Focusing, SGP success and failure

In a nutshell, I am wondering : How does SGP determine success and failure in autofocusing?

========================
I have just figured out how to calculate the Step Size and posted in another thread. While I was figuring that out, I thought about auto focusing in general:

The [SGP’s] expectation is that the user will bring the imaging system into a focus condition that is acceptable to the user. That focus condition is characterized by an HFR (Half Flux Radius) value and is related to the focuser mechanism being at a position P1.

P1 left unchanged, and if nothing changes in between us and the target throughout the night, HFR1 should remain the same. But that is not realistic. Things do change throughout an imaging session - things like temperature, humidity as well as other atmospheric effects.

If we start by achieving [what we accept as] a good focus through whatever means that we can use (Bahtinov Mask, etc), and if we bring the system out of focus by an HFR = 3 to 5 times HFR1, and select a number of data points to be used “when auto focusing is activated”, then everything should work out fine: SGP knows what HFR to shoot for, it knows how many steps the user wants focusing data to be gathered in, and how far out (and in) it needs to go to begin searching for that HFR1 again.

How far out of realistic is this frame of thought?

Farzad

I’m not exactly sure what you are getting at here.

This is 100% true. It is the only reason auto focus exists… It is designed to correct for precisely this.

I was only saying that “I got it”, and wanted to share with an unsuspecting visitor.