SUMMARY When using focus offsets, guiding may be disrupted at the beginning of the first exposure after focus or align, resulting in worse than normal guiding performance. STEPS TO REPRODUCE 1. Have a two optical trains setup with unfiltered OAG, filter wheel in the main train, single focusser in both trains. Add R and O filters with different offsets. Lock O filter to R for Focus. Set Align to use R as well. Have a repeating job in the schedule with Focus, Align, Track, Guide enabled and sequence with O exposures. 2. Start the schedule. First iteration may or may not hit the problem. Subsequent ones will as they have just ended with O filter selected. 3. Focus switched the wheel to R, completes, switches to O and applies offset. 4. Align switches to R again and applies the offset (a bit excessive work but fine). Runs successfully. 5. Guide starts (main train is still on R filter!), captures a couple images, locks on a star. 6. CCD starts its sequence, selects the O filter, suspends guiding for focusser action, applies the offset, resumes guiding. 7. Guide sees its star moved a bit by focusser shake and starts correcting. OBSERVED RESULT With my mount and focusser, likely, that shake is below configured tolerance, so exposure continues but the image is worse than it could be. Guide may take multiple iterations to move the star back to where it wants, especially if significant Dec correction is required and that hits backlash. I often see 3" swings in Dec over first 30s of a 2m exposure. That yields like 2×4" stars. I set guide tolerance to 2" which is already quite strict for my weather. EXPECTED RESULT Multiple options could work better. Perhaps, restart guiding after moving focus rather than suspend and resume. IOW, force Guide to lock on star in new position. To me, this sounds much like what happens after dithering, so maybe reuse that logic somehow. SOFTWARE/OS VERSIONS Ubuntu 24.04.1 KStars 3.7.5 from PPA
Konstantin, I would think that if one is guiding with a OAG or an ONAG, where the guide camera is behind the focuser, then it is normal for amateur equipment to see some shift in the image (and thus in the guide image too) when the focus point moves. This can be movement in focus just from auto-focus, or from filter changes. Thus, I don't think this is a software issue, just physics/optics. The solution I believe is to use the capture "limit" setting "Only capture if guide deviation < ___". Then after auto-focus is done, or when the filter-caused focuser change is done, and when capture starts back up, capture will notice a large guider error and not re-start capturing until the guide error settles down.