Auto-focus should be more resilient, not accept worse results that at original focus position

I had auto-focus fail in the middle of a sequence - it introduced bloated stars until the next auto-focus round, which restored everything.

Many things can throw auto-focus off, from increased LP, wind, vibrations, mount performance, to clouds.

Multiple things could be done:

More resiliency: 1. Introduce a number of exposures per autofocus step setting N, defaulting to 1 2. For each autofocus step, take N exposures 3. HFR for that step is the average 4. proceed as usual

Recover from bad auto-focus: 1. Record original focuser position 2. Take an exposure (or multiple, as per the resiliency piece) at original focuser position, record HFR / average HFR 3. Proceed with autofocus as usual 4. Take exposure (or set of exposures) at end of autofocus, get HFR / average HFR 5. If final HFR is more than 1.15 * original focuser position -> return to original focuser position.

Plus

Introduce a restart of the auto-focus, with a "Max number of auto-focus attempts" setting, defaulting to 1.

Comments (2)