I don't understand the interactions among focus triggers in image *sequences* (Camera tab) and in the *scheduler* (scheduler tab).

Tonight I got into what seems like an infinite loop of autofocus (no captures!).

WHAT I WANT TO DO:
* alternate 10min guided Ha and O3 captures all night
* dither every other frame (after each pair Ha/O3)
* autofocus once an hour using L

HOW I TRIED TO DO THIS:
* Created a sequence queue (.esq) with one 600s Ha and one 600s O3 frame
* Check the box under "Limit Settings" that says "Refocus every" 60 minutes
* Created a scheduler list (.esl) that calls the sequence queue with "Repeat for 20 runs"
* Check all four "Steps:" Track, Focus, Align, Guide in the Scheduler List

WHAT WENT WRONG:
* Autofocus mania!
* Scheduler starts up
* Slew and align successful
* Autofocus successful
* Start guiding
* Another autofocus starts
* First capture in .esq (Ha) is aborted because autofocus is in progress
* Autofocus successful
* Another autofocus starts
etc etc etc ...

I've attached a relevant log.

What's really maddening is that after this I tried everything I could to stop the crazy autofocus demon:
* Unchecked the "Focus" step in the Scheduler. Nope
* Unchecked the "Refocus every 60 minutes" box under "Limit Settings" in the Capture module. Nope
* Stopped running the scheduler completely and just ran the sequence from the Capture module. Nope.
* Disconnected the focuser completely from the INDI Control Panel. Nope. Even then the sequence just waits forever for the nonexistent focuser!

For each of these attempts, I completely disconnected the equipment, closed Ekos, and quit KStars. Then restarted.

The zombie autofocus loop just continues, and no frames are ever captured.

Finally, I created a new INDI equipment profile that doesn't even *have* a focuser, and finally I can image. Of course my images won't stay focused all night but I'd rather collect less-than-perfectly focused data during the new moon than no data at all.

Can somebody help me to set this up correctly so that I can proceed PAST autofocus and get on with capturing images?

Thanks!
Scott

Really the same issue, after update of my Stellarmate.

Autofocus after guiding procedure. In linear mode, the focus move to +200 steps, and the autofocus procedure seems to be blocked.
I tried without the focus (on schedule), but it's the align module that seems to be blocked.
Kstar crashes if stop the procedure (focus or align).

Scott, Supaii,
could you please be so kind and turn on the verbose mode - at least for focusing and capturing. The behavior is strange, but it is not clear what happens.

Cheers
Wolfgang

The star finding algorithm seems to get confused when the frame is too far out of focus. I run into this problem when switching filters that are not para-focal. If the new filter is far enough away from the focus point, the first step of the re-focus will not back up enough, but if it backs up too much the a refocus of an existing filter will fail. The biggest problem seems to be that the algorithm for finding stars will trigger on the edge of an out-of focus star instead of the whole large round disc.

One other thought on this: is the filter offset used at all by the focuser as a starting point based on the currently focused filter? If not, can it be? Once the offsets are set up, this seems reasonable.

Jim

Jim, I've had that happen too, but that's not what's going on here. I'm able to get bang-on focus very quickly every time.

The trouble is that focus gets triggered over and over (successful each time) and the scheduler never moves on to actual imaging. It's just all wonderful successful autofocus all night (or rather until I kill it).

Wolfgang, it's daytime here. I will try to reproduce with simulators and save verbose logs.

Thanks,
Scott

Ah, the thing that comes to mind then is the refocus if HFR > # setting on the CCD tab. How did that number compare to the actual focus you were achieving?

Jim

Wolfgang,

Well drat! The endless autofocus problem does NOT occur in the simulators.

I'm stumped. I will wait until dark (about 8 hours from now) and try again with verbose logs.

Thanks for looking!
Scott

Jim,

Thanks for this helpful suggestion. I didn't have that box in the Capture module ticked, although I *did* select "Refocus every 60 minutes."

Once I got stuck in the infinite autofocus loop I tried unchecking that box in the Capture module, then unchecking the "Focus" step in the Scheduler,then stopped the Scheduler completely and just running a sequence in the Capture model, then disconnecting the focuser in the INDI control panel. With each attempt I completely disconnected and stopped INDI and also quit completely out of Ekos and KStars, then restarted.

The only way I was able to collect data at all was by creating a new equipment profile in INDI that doesn't *have* a focuser.

I've attached a log in case that's helpful, and plan to repeat with verbose logs when it gets dark here in about 8 hours. No joy from the simulators.

Thanks again for your help!
Scott

Scott,
could you please take a closer look whether you have a setup for refocusing on filter change? Maybe in combination with defining another filter for focusing you are creating an endless loop?

At least I can see in your log file that after the successful autofocus a filter change happens which triggers focusing and this seems to lead to another filter change.

HTH
Wolfgang

Wolfgang,

THANK YOU -- I think that was it!

I had misunderstood the meaning of the "autofocus" checkboxes in the Filter Settings panel.

What I INTEND to do is to autofocus ONLY with L and then use filter offsets for all other filters. But I had checked the "autofocus" boxes in the Filter Settings, so it was trying to autofocus very filter change. Now I've unchecked all the "AutoFocus" boxes and instead set them all to Lock Filter = L.

I won't know for sure until I can test it tonight, but THANK YOU for looking carefully at this. I'm sorry to have wasted your time.

Scott

SOLVED: Checking the "AutoFocus" tick boxes in the Filter Settings panel produces an infinite loop when filters are changed.

Thank you Wolfgang for figuring this out for me.

Scott

Happy to hear!

Clear skies
Wolfgang