Hi,

I've seen it already several times, but wasn't sure if I did something weird. For sure I didn't yesterday. This is what happened:
I had created a sequence in the camera tab of EKOS (+ button) that was supposed to take 120 frames of 30s each.
After a while I thought I'll go for a refocus and wanted to pause the sequence. So I pressed the pause button below the sequence table display. What now happened perplexed me:
It would finish the current exposure.
Then the activity display (text next to the exposure) told me 'Aligning'. From the beep I gather it did a (successful) plate solve. Why? After that, it <strong>continued</strong> to take frames.
The two buttons (stop and pause during a running sequence) had changed to play and pause, with pause greyed out and inactive. Pressing play <strong>aborted</strong> the sequence.
Looking at the log file, after pausing it executed a meridian flip! The target (Coat Hanger Asterism) was many hours before being even close to the meridian!? (rise 20:50, transit 3:37, time of the event 22:32). I had not even activated any automatic meridian flip activity... What's going on there?

Hi,
it looks like you stepped into a very special situation. You paused a sequence exactly at that time when the mount decided it was time to execute a meridian flip. Assuming that this was correct, the behavior in general is right:

1. Wait until the sequence is finished, then pause and do the meridian flip.
2. Re-align after the meridian flip - that's why you see "Aligning" in the Info tab.

What should not happen is that a paused sequence restarts automatically after a meridian flip. Can you reproduce this with Simulators?

Next point: EKOS executes only a meridian flip if the meridian flip is activated on the Mount tab. Are you really sure that the meridian flip on the Mount tab is deactivated?

So let's assume it was checked. But why did it start so early? As far as I understand, it was roughly 3h too early. So there must be something wrong with the location or the KStars time. Because the meridian flip is only executed when the HA of an object changes from a negative to a positive value.

Hope that helps
- Wolfgang

It was less than two hours after rise, more than 4 hours before meridian.....

Who decides to flip? The mount, or EKOS? I thought (only) the latter.

I can try - I haven't used simulators so far. But of course, for an automatic meridian flip the sequence should continue after the flip....

Quite sure, yes. It wouldn't be too helpful for the CEM60, as the iEQ driver unfortunately doesn't support pier side info, so guiding would go crazy without a (manual) calibration flip in PHD2.

Indeed. Even if the option had been set, it would have been for HA zero. If, for whatever reason, the pointing was giving wrong info on the HA - why dit GOTO to the target work perfectly?

What I could imagine is that pressing pause does a check if a flip is needed, and that check goes wrong, e.g., because the pier side info isn't there?

Depends on your setting, your mount and the driver you use. Here you can find an explanation how the meridian flip works: indilib.org/forum/general/5296-mount-doe...dian-flip.html#40069

Because EKOS issued a re-alignment after the flip was completed. Did the pier side change?

I’ll check it with the simulators and pausing.

OK, problem confirmed, this is a bug. No matter where you are pointing to, the Capture module claims that a meridian flip was taking place when it paused.

- Wolfgang

Problem solved, I submitted a fix: phabricator.kde.org/D22013

Thanks so much!
Just compiled it, will test ASAP