Plate solving problems

[mcamilleri]

I have been having endless problems with plate solving and PA. It seldom successfully plate solves, and occasionally I seem to find the magic combination and successfully sync the mount and it works for a little while, then fails. The usual error is reporting it can only see 20 stars, even though on screen I can visually see dozens. I have tried exposures up to 10s, various gain settings etc and it makes no difference.

I have ASTAP installed, H18 database. Images from the same camera/scope successfully plate solve using ASTAP in NINA with 2-5s exposures or using ASIAir.

I have a CGX mount, running through CPWI to Sharpcap. The mount operates normally from Sharpcap and the mount position is showing. Mount is either polar aligned already or to within a couple of degrees.

Cameras are either asi485mc or Orion G16. Scope is C9.25 at either 785mm or 1440mm focal length (3.3 or 6.3 reducers). I input those focal lengths into the plate solving settings.

I have tried various camera configurations, RGB24, RAW16, MONO8 etc. At one point it seemed that it worked better with RAW but that didn't persist.

Attached is a log immediately after a plate solve failure, with the image (admittedly a pretty crappy one, with too large a sensor for the 3.3 reducer causing vignetting and distortion, and poor focus, but NINA and ASIAir usually solves similar images). The image is jpg as png was too large to load

I expect I am just doing something stupid or have failed to specify a key setting correctly, but I have no idea what that might be.

Kind regards
Michael

[admin]

Hi,

that's odd - I just loaded your image into SharpCap's 'Test Camera 1' as a test image and then hit the plate solve button and it worked in Astap using the V17 star database - it also found 55 stars in the image.

It could be that you have a different version of Astap to me (I have one from Jan 2022 at the moment, but I know that some older ones had issues with the search not being quite comprehensive). Alternatively it could be some settings in Astap - here are my plate solving settings in Astap - maybe the 'slow' option helps?

Capture.JPG (41.43 KiB) Viewed 609 times

Anyway, worth checking these things in Astap before anything else, as it looks like it should work.

cheers,

Robin

[mcamilleri]

Thanks. I'll take a look at the ASTAP settings, which I have not looked at or changed yet.

I have a fresh install of ASCOM and ASTAP as at a few weeks ago, and am using the H18 database as I need to solve down to a 0.2 deg FOV. I did have ASCOM and ASTAP installed previously but did a full uninstall.

I don't have the H17 or V17 database loaded but my understanding is that I don't need them unless I want to solve wide field images.

Michael

[admin]

Hi Michael,

yes, I'd agree, in fact the H18 should work across a wider range of scales than H17, but may be slower as there are more patterns to check.

It may be worth loading your image into Astap itself and playing around there to see if you can persuade Astap to solve it directly - easier to tinker with the parameters in Astap than to keep starting Astap, changing something, closing it, trying SharpCap, etc.

cheers,

Robin

[mcamilleri]

Hi Robin,

I tried that image using the SHarpcap test camera and it failed on a blind plate solve. It also failed in NINA on a blind solve.

I tried ASTAP and it failed on a blind solve at 180 deg, or when given the RA/DEC from Astronomy.net.

It did solve in ASTAP when manually set the FOV to 0.46 deg ( the calculated image height). It fails when the Astap pre-defined 0.5 deg FOV height is used.

I tried specifying the command line setting in the Sharpcap plate solving settings "-fov 0.46" and that failed.

So perhaps it is taking the image FOV and rounding up to 0.5 deg and it then fails.

[admin]

Hi,

that's odd, since when my Astap succeeds in solving, it does so at a scale of 0.56 degrees, so that's even further out from the real scale than 0.5 that fails for you.

The things I'm thinking about now are

* Some difference between the H17 database that I'm using and the H18 one that you are using is causing different scales to be tried or different results
* Some difference between my Astap (about a year old) and yours - perhaps a new issue introduced.

Does it work any better if you don't have to do a whole sky search? You can tweak this in SharpCap as follows

* Set up SharpCap to talk to the ASCOM Telescope Simulator in the hardware settings under 'Mount'.
* Run the deep sky test camera
* Connect the Telescope simulator driver and slew to RA=~1h and Dec=~-75
* Press the 'plate solve and sync' button in the menu or the equivalent button on the telescope controls in SharpCap

This process will run the search only in the 15 degrees around the current mount position, which may give better results. There have been some issues in Astap previously where it didn't quite do full coverage of all the possible sky positions, particularly for areas a long way from the starting point. I thought these were fixed, but you never know...

cheers,

Robin

[mcamilleri]

For some reason ASTAP has 20 stars max as the default. I increased it to 200 to match your settings, it didn't seem to make a difference to the FOV behaviour. I tried the max stars "-s 200" command line setting from Sharpcap and that didn't make a difference.

I tried various search ranges from 15 to 180 deg in ASTAP and that didn't make a difference.

I'll download the H17 database and try that.

Michael

[mcamilleri]

ON the ASTAP site it says that the databases will be phased out and new ones phased in. At the moment the H17 and H18 databases for Windows are now deprecated and replace by the d05, d20 and d50 databases.

https://www.hnsky.org/astap.htm

[mcamilleri]

So I downloaded d05 and d20 databases, uninstalled h18 and ASTAP complains about h17/h18 not installed. It seems ASTAP may need to be reinstalled after uninstalling and changing to the new databases.

[mcamilleri]

Plate solving seems to be much happier with the d05 and d20 and reinstalled ASTAP. It plate solved my two test images first time, either with simulated mount position or blind. So hopefully it will work for real tonight.

Thanks for your help

Michael