Astrometry.NET doesn't work if ASTAP is your main solver

Hey it seems to me that if you have ASTAP configured as your main solver, Astrometry.NET won’t solve either in a blind fail over OR as Solve and Sync Blind.

While waiting for SGP/Astrometry.NET to complete I was able to load the same fits file to nova.astrometry.net and get a successful solve, so aborted the SGP solve.

Was not able to get any successful blind solve until I went back to PlateSovle2 as the main solver and then things returned to normal.

I had seen another post here (which I can’t find now) about blind failover not working with ASTAP, and I wanted to add that it isn’t just the fail over that not working. It’s the blind solver in general.

I’m not sure that I’ve seen this behavior and I use this combination. I’ll run some tests and attempt to duplicate.

Jared

Maybe something odd with my sub exposures from this latest setup. As PlateSolve2 seems to have problems with them even after a entering the correct image scale in the control panel (by successful blind solve). Maybe due to some field curvature? Not sure what to do about it as I don’t think we have any ability to say, only look at an ROI for plate solving, or other parameters?

Example sub, if you have time to look:

ASTAP can solve it.

I tried forcing a blind fail over from ASTAP by mis-configuring the imaging scale but solving with ASTAP from the image context menu still worked.

I’ll try tonight again with ASTAP as the solver and see what happens.

Thanks!

I opened it and solved with ANSVR no problem: image

You can right click the image, click “Plate Solve” and use the “Blind Solve” option so you don’t have to mess with things to invoke the fail over.
image

ASTAP also solved it…so not really sure what the issue could be.

Jared

Hi.
I have the same issue each time. No blind solve if the main solver is Astap.
JM

I repeated the experiment, with the posted image, on a second computer.

With PlateSolve2 configured as the solver, Astrometry.net set as the blind solver, and the image scale set to 1.34 "/pix in the control panel:

  1. Right click on image and select solve. No solution after 900 regions. Hint data is correct.

  2. Right click on image and select blind solve, solution reached after about 3 minutes.

Configure ASTAP as the solver.

  1. Right click on image and select solve. Immediate solution.

  2. Right click on image and select blind solve. No solution after 800 seconds.

Just as I was typing that last sentence I saw a dialog flash and disappear but the status at the bottom of SGP still indicated that the blind solve was proceeding with the counter still counting. I checked underneath the main SGP window and a dialog was there saying it solved.

So it did eventually blind solve, although it took much longer, and the dialog was hidden without the status changing until I clicked OK in the hidden dialog.

I would have missed the dialog if I had blinked or been looking away.

Are you using the online version of Astrometry.net or ANSVR locally? I’ve found the online one to be all over the place with how long solves take. Generally it’s pretty fast but it seems to depend on load a lot. I typically find ANSVR to be substantially faster and more reliable than the online version.

Jared

I tend to use the online version from home, and ansvr in the field, but these days I’m always home.

I can test with ansvr and report back.

Was this related to my thread? I’m not sure if Jared had a chance to test this ?

Thanks

I’m not sure if that’s the one I saw or another.

There seems to be a few problems.

On my computer in the house, with ansvr as the blind solver (and ASTAP as the main solver) ansvr solved the image in 13 seconds with dialogs acting normal.

I will configure my telescope computer to use ASTAP and ansvr and see what happens.

25 seconds to solve with ansvr on the telescope computer (ASTAP also working). So I will roll with that setup.

Thanks,
Glenn

I replied on your thread. But I think these may be somewhat related. Basically the status bar isn’t updating correctly when blind solve failover is invoked. So it appears the primary solve is still running when it’s actually running blind solve.

Jared