Just to give some background, I’ve always used PS2 for plate solving which has worked well except for wider angle images (image scales larger than about 5"/px). PS2 just doesn’t work on those in 95% of cases. So after some research ASTAP was highly recommended. So I downloaded ASTAP, both their H18 and W08 databases. I changed the settings in SGP to point to the ASTAP directory and all was well, easy enough. My first plate solve with a 135mm lens (10"/px) solved in about a quarter second and appeared to be centering the scope accurately and working well. However, when it came time to manually rotate my camera, I quickly found out that ASTAP generally had no idea what to rotate it to. The initial rotate instruction seemed to be about right. But after I rotated it to approximately that angle and solved it again, it now said I had to rotate it an additional 48 degrees or something similar. Obviously this wasn’t correct. I tried it again. This time it said it had to be rotated by 30 degrees or thereabouts. Then 6 degrees. Then once it said I needed to rotated it clockwise, so I rotated it a bit clockwise, then after solving it it said I needed to rotate it clockwise at a larger angle than I had previously. Basically, it had no idea what to rotate it to but seemed to center the mount well enough. That was issue #1.
Issue #2 was far worse. I was doing a 2-panel mosaic near Pleaides and around 00:15 had a meridian flip. The meridian flip went fine and the solver placed the scope accurately. But about 35 minutes later when it came time to go from one mosaic panel to the next, the plate solve somehow sent the mount nearly 40 degrees closer to the western horizon than it should have. The odd thing is, the log of the plate solve seems to think that it was in the correct location (which would have been roughly RA 3:56 and Dec 35) but it wasn’t anywhere close. So I had 50 images of the edge of a tree and the back of my 4runner. Always fun to wake up to. So I have absolutely no idea how that could have happened. The log is in the link below. Anyone have any ideas how this could have happened?
Does it seem like you are doing the exact opposite of what is required or are you saying that there is no pattern? Certain rigs, will require that the directions for rotation are reversed based on how the final projection onto the sensor lands (mirror flipping, diagonals, etc). There is an option on the “Other” to reverse the corrections.
Not sure about this one. Unfortunately SGPro’s visibility for slewing ends at the boundary between it and the driver. If this looks right to you, then it’s fair to say that it likely is not an issue with SGPro or the solver (not a guarantee…). This is what we know:
Post flip centering begins
Solved and synced to JNOW RA: 4.01044043645123 Dec: 34.7811666893031
Slewed to JNOW RA: 4.00309994635801 Dec: 34.7792865840104
It’s hard to say why the scope would go the wrong position here, but it’s likely not a coincidence that the issue occurred after the flip. That said, the telescope is reporting that it thinks it’s on the correct side of the pier. Have you made sure to update to your scope’s latest driver (or downgrade if you upgraded recently)?
Re #1, it seemed to be a mix. There were times where it was completely random and at least twice I thought it was close on the angle but the wrong direction. Also of note, the rotation symbol always pointed clockwise (I think) despite SGP saying it needed to be rotated counterclockwise. I thought that was odd as I don’t remember it doing that before. In any event, I had never had an issue with it displaying the correct angle or rotation direction with PS2 prior to this, when it actually solved.
Re #2, I have the latest drivers and everything should be up to date. I did just get the Dec board on my mount replaced a couple weeks ago, but everything appeared to be working fine with the mount prior to this mosaic panel slew, so I doubt it was that. I wish the log was able to differentiate if the slew to the wrong location happened prior to the solve or afterwards, as that would indicate which was likely at fault. Regardless, the solver should have discovered it was in completely the wrong location and failed the solve, rather than thinking it had solved correctly and going about its business. That’s what leads me to believe it was the solver.
Related to ASTAP, since I downloaded two different directories, is there anywhere where I have to indicate which directory to use or does it choose automatically based on my image scale setting?