So… I’ve been having great success, lately with SGP, (and a week straight of clear skies)… tonight I’m trying to catch M76, the Little Dumbell Nebula with Narrow Band Filters and a 2X Barlow Lens. My issue is that it won’t plate solve the initial image in order to center, and so the sequence fails. I’ve tried twice.
I’m still using the last GA Release: 2.5.1.17
First: I am willing to accept I’m doing something stupid; I’m just hoping someone can tell me what that is.
Second: When I right click on the photo that was taken for the initial center and sync, (the one that SGP just told me, after minutes of waiting, could not be solved, even after failing over to blind solving), and ask for a blind solve, it returns a reasonable set of coordinates within mere moments.
Ordinarily plate solves work well. I use the local astrometery server. Also, I ussualy do not use a Barlow lens.
well with the barlow in place your plate scale has changed… the blind solver does not care about this but PS2 does and quite possibly when astrometry is invoked after the failure of PS2 SGP has passed the plate scale.
so… did you update the plate scale in the SGP config when you inserted the barlow?
i don’t think so, but i only ever have to change it very seldomly, so i might have forgotten if there is another place. i started thinking about star quality and failure to identify enough stars but it does not make sense that the local ansvr server would work in one situation and not in the other, unless for some reason the --sigma value is different in each invocation, and i would not exect that.
How are you determining the new plate scale? Calculation is unlikely to be accurate enough because the barlow magnification factor varies with the distance from the lens to the focal plane. Also if you are using a SCT the SCT focal length will vary with the focus position and a barlow usually needs a considerable change in focus position.
IMO the only reliable way to get a good plate scale is to solve the image using a blind solver, i.e. Astronomy.Net. Do this with the exact optical setup that you want to use for your imaging.
I’m sure you’re correct that I’ve used a too naive calculation for the image scale… but… that leaves open the question why when I manually asked for a blind plate solve it worked more-or-less instantly, from within SGP, (wrong-click on the image, select “Plate Solve”, select “Blind Solve”), but failed repeatedly when SGP was trying to do the initial centering, (and reported that it had failed over to a blind solve.)
In any case, I think, probably, this should be abandoned. Continuing to work with SGP last evening, all sorts of odd things started happening with the profile I was using… at one point, the camera definition vanished. I’m unsure what I’d done to cause that, but likely the plate solving issues was related, because when I started looking, there were strange things going on in the profile. (It had, for example, lost track of the fact that I use the local asrometery solver.) I’m going to just crush the equipment profile and rebuild it. I’ll see if that doesn’t solve this.
also if starting over with a fresh profile does not help, i think a “global reset” command line option was added to the 2.5 betas. there is a thread here somewhere about it.
I’ve been contemplating hoping over to the beta release. But I had been thinking that I should avoid introducing new variables.
After plate solving went awry, the next bizarre issue was that SGP started doing autofocus on every frame. I think that revolved around the setting autofocus on filter change, and even though the filter was the same from one sub to the next, SGP issued a change filter to the same filter ahead of each sub. I have no idea what that was about. I can only assume that I’ve foolishly broken something, but damned if I can guess what.
well for what it’s worth, the betas have been pretty stable for me. i have discovered/reported some bugs (which have been fixed) but each of them had a workaround. IMO the beta is pretty solid but i understand your point of view - i would have liked to be able to install betas alongside release versions but alas it does not work; the installer always blows away the old installation, no matter where it is or whether or not it is beta or release.