Hi everyone
After a meridian flp, the solved image is over 500px out of alignment; really annoying!

• Set RA and DEC in scheduler: solves and captures fine
  perform the flip
  align does not solve to the RA and DEC entered in the scheduler

What am I doing wrong?
Cheers and TIA for any pointers.
logs: drive.google.com/open?id=1abiDjaFICEPQk6F_LbzROGPILKT8XN0J

This is a mosaic right?

From the log, I see that the flip occurs at 00:42, while the scope is looking at 20:28:07/40:28:24.

While scope is on west side, alignment accuracy is 18 arcsecs, on east side alignment accuracy is 6 arcsecs. Well, 12 arcsecs probably don't translate to 500 pixels in your field of view, do they?

I think I see you have alignment calibration reset when processing a flip, so OK.

I also checked PHD2 logs, but it just selected and locked a star and started guiding, nothing else, for 14 seconds, be it before or after flip... I don't know PHD2 very well, but I don't suppose it could move the guiding star back to the original position? So that would defeat the post-flip alignment? That's probably completely stupid, I can't verify where your guide star is located precisely nor if the same one is used before the flip and after the flip. Pre-flip is X=454 Y=100, post-flip is X=157 Y=55.


-Eric

No, not a mosic. It is a 812mm aperture, f3.9 reflector.
The guiding is fine; PHD2 finds a guide star in the FOV it's given. It never changes the fov.

I got the coordinates by solving the this image in align and then transferring those coordinates to the scheduler. The fits header contains: RA 20 26 46.15 DEC 40 20 50.54. The fits file is attached. Maybe align solved it wrongly?
Any ideas why it's out by 500px?
Cheers and thanks for your patience.
Steve

Ah I think I misunderstood your startools screenshot. You solved the original image (sorry I don't have a fits viewer on my phone), and it gave you 20:26:56/40:20:50. In the log, the target that is aligned does not match these coordinates. The mount aligns towards 20:28:07/40:28:24. That coordinates offset is probably the source of the pixel offset.

I think you got bitten by the RA/DEC bug I fixed recently in the Scheduler UI, in which double-clicking the job item, updating the RA/DEC and saving immediately does not actually save the target coordinates properly. Check that your scheduler job list is providing proper RA/DEC again, my guess is that it doesn't.

Reference: phabricator.kde.org/D14679, two weeks ago.

-Eric

Ah, OK. has this reached the nightly builds?
Thanks.

Yes it did. Which version are you using?

-Eric

nightly 6 Agosto:
-rwxr-xr-x 1 root root 10468544 Aug 6 20:03 kstars

You should provide what is in the about window instead. Please check your coordinates in your .esl file too, to confirm. I assumed you did enter the right RA/DEC but maybe I'm mistaken.

The log (log_23-55-04) indicates a first (manual?) alignment to 20:27:25/40:24:42 before the scheduler is started. Then when the scheduler starts, the mount is requested to slew to 20:28:04/40:28:21.


-Eric

...about window?
Yeah, it's a (my) mess as I didn't save the scheduler file. It seems that the coordinates entered in the scheduler are not those passed to align:(
Going again now. This time I've saved the esl but already I can see different coordinates have been used in align.
Logs to follow.
Thanks for your time,
Steve

Help -> About KStars

Thanks:

Build: 2018-08-06T18:03:59Z

Hi. Here is the log from the last session. You can see the RA:DEC discrepancies in the attachments.
Would it be possible to align to the same coordinates everywhere? That would be great.
TIA and clear skies.

logs: drive.google.com/open?id=1kxP6HE5XcaD0hfWRvs8ettKOaEhy2dCd