I use Frame and Mosaic wizard to create a sequence. All is well and the centreing takes place as required and gets the scope in the right place. Then I come to a meridian flip… It plate solves and says that the solve is correct but when the frame comes in it is clearly not in the right place.
I am using a Mesu and have the telescope set to Sync. I have attached the SGP log (the meridian flip is at the end of the log) as well as the 2 images (in jpeg) so that the misalignment can be seen.
I have just been out and watched the mount do another meridian flip on a different target and exactly the same has happened. The mount is not moving in anyway and I’m pretty sure this has only started happening since going to the latest stable veriosn.
I have also experienced this and, like Sara, think it is a new issue. The meridian flip used to work perfectly. I have had at least two runs when the post-flip image was well out. Not sure why this is now happening.
[Edit: I’ve just checked Sara’s image and it turns out we are imaging the same object!! This is probably not relevant, but I mention it anyway.]
I have also tried this on other targets last night Steve - So in this case it is NOT target dependent. I hope it can be sorted as it at the moment an automated meridian flip is just not possible.
What version of SGP are you using Sara? I suspect that this is not a Sitech/Mesu issue but a SGP issue. It is interesting that one of the changes in v18 was “Reintroduced the initial solve process to the front of the auto-meridian flip routine (notice there is no attempt to sync here) to get a proper post-flip reference. This should address a problem that could cause subsequent centering routines to be "off"” I wonder if it is this we are suffering from.
I have the latest version installed and have experienced this problem. It was occurring (for me) with previous version as well. I was imaging M81 and M82, the pier flip happened, and one of two things would occur… 1) the plate solve after pier flip would indicate fail (3 times) yet the galaxies were in the FOV. 2) The plate solve indicated success but the galaxies were outside of the FOV.
Looking at the log the post flip centring process finishes at 22:49:21.972 when the error is 2.2 pixels from the target position of J2000 RA: 9.91578194949421 (09h54m56.82s) Dec: 69.0907793888959 (69°05’26.81").
This seems to show that the mount reaches the target position, however from a look at the images this is different to the position before pier flip.
There is no image collection and solve before the pier flip.
The sequence started at 21:23:22 and the target position is J2000 RA: 9.92325 (09h55m23.70s) Dec: 69.1130027777778 (69°06’46.81") and it is set to this position to within 3.3 pixels.
So it looks as if the centring process on both sides of the meridian is fine and the reason for the difference is that the two positions are different.
The question seems to be where does the post flip position come from? I can’t see any place in the log where the RA of 9.91578 appears until the start of the post flip centring process. The pre flip position seems to come from the sequence.
Has the post flip position come from the mount position before the flip? While this was synced to the target position at the start of the sequence mount errors and guiding mean that it won’t necessarily be the same position when the flip is commanded, in this case over an hour later.
It looks as if either the sequence target needs to be used or the position on the sky determined immediately before the flip by collecting an image and solving it.
You may want to grab the beta. This has been “fixed”. Essentially we had removed the plate solve prior to flip thinking that we could trust the position from your scope. This has proven to be a bad assumption.
We added the plate solve back into the beta (in 2.6.0.18 specifically) which solves this issue. There is no sync done, only a solve to get the reference location prior to the flip.
Thanks Jared… I don’t normally run with Beta’s as I’m not one for problem solving, but this was driving me mad, so I’ve given it a go… hope it’s a fairly stable Beta
