So, I decided to plunder forth and do some testing. I used Deneb as my target, and locked on with PHD2.
My new EAF focuser is working better and better. Last clear skies, I wanted to see how my filters stacked up for what I call "Repeat accuracy". So I chose a filter, LUM, first since it is in Pos. 1, and got the Auto focus in focus. I got 19707 steps as the result. OK, my baseline.
Then I changed to RED, Pos. 2, and ran the Auto Focus with it. Got 19711 steps as that result. I was pleased, the two filters were really close in focus.
So, I forged ahead collecting results through my filters. 19711 steps were the highest out of 6 filters, 19700 was the lowest out of my filters.
The average was 19707.166666> steps. (19707 steps).
So, I'm satisfied my filters are all relatively very close to being in focus as the different filters come into play. I think I am pleased with the matching ZWO ASI did with the set.
Granted, this is just a shade tree mechanic test in the back yard. Not really of value to anyone but me.
I can zero out my focuser at the start (or any time) and just start over. But since it is a Crawford Styled focuser, it has no teeth and is just a friction drive.
If I ever get a new Telescope, I will be sure it has a Rack and Pinion focuser. And I will inquire about backlash as well. But for now, I'm stuck with a Crawford drive. It is what it is.
(But the Repeat Accuracy simply isn't there with this focuser. In my honest opinion.)
So Zero is a bit ambiguous. Best bet is to zero the focuser after alignment near the intended subject for the session, then do the series on the chosen filter. Change to the next filter, run auto focus, and run the next pile of images.
Are you guys doing something like this?