ASTAP unreliability with widefield images

[Kelvin Hennessy]

Hi. For several years I have tried different platesolvers with widefield images (most commonly 85mm & 135mm on a full frame Canon 6D) and never had any luck getting any to work reliably. I thought I would have another go at ASTAP today but once again been an exercise in frustration.
Firstly the latest non-dev version wont recognize the D50 database is installed on Windows 11. After much searching I found a link suggesting to try an older ASTAP.EXE version. I went for the latest dev version instead and it works. That's not a good start !
I loaded one of my FITS images and tried both blind solving and setting coordinates close to the exact centre in as a hint but neither would solve. I've run the star tests and they pass. The quads test fails with a "floating point division by zero" error most times but sometimes works and shows a bunch of quadrilaterals.
I've attached the astrometry.net solved image to show that it can be solved.
I have tried both D50 and G05 databases

Am I wasting my time with ASTAP on such wide field images ?

thanks, Kelvin

[Kelvin Hennessy]

Well I've given up on this once again. I dont have time to waste on faulty software. I have tried downloading a few older versions and keep getting the attached error about not finding a database. I have tried installing to c:\astap instead of the programfiles folder without success. Only the latest dev version could find the database. This is a really bad look for your software to have such a persistent problem !!

[han.k]

Hello Kelvin,

I managed to download your image from nova.astrometry.net and test it. Solving is no problem. There are some difficulties because it is undersampled so stars are pretty tiny. This could have potential problem if the image also contains hot pixels. Then binning SHOULD be set at 1. Otherwise the small stars can be ignored.

Your error.png show ASTAP is located in c:\astap. That is no problem but you should also install the databases there. All files are normally installed at c:\program files\astap. You can install them anywhere even manually but you have to keep the astap.exe executable at the same directory as the databases. So move all files or c:\astap or c:\program files\astap. But probably it is better to unstall all and start fresh and let the installer do the work.

Looking to your image, the field is 28x19 degrees , install the G05 database. You can remove/uninstall all other databases.

Look to the screenshot and settings.
Set max stars at 1000.
Binning at 1
Field of view (height)=18.96 or 19 degrees
Radius search area=180 degrees

Load an image and the settings: maxstars=1000 and binning=1 and hit either the solve image button to save these settings or go to viewer, FILE, SETTINGS, SAVE SETTINGS.

The field of view should normally be provided by SGP. With the current specified pixel 4.63 micrometer ASTAP report this:

Set FOV=18.96d or scale=24.2"/pix or FL=39mm

So you have to enter in SGP about 24 arcsec/pixel. Focal length is calculated 39 mm. If you change the focal length you have to adapt.

So do the following:
1) Install ASTAP correctly including G05.
2) Load one of your images
3) Enter the settings for screenshot
4) Check if they solve
5) Adapt in SGP the settings for pixel size of 24"/pixel

If it still fails, ask again.

Han

[Kelvin Hennessy]

Hi Han
You instructions have worked well for me and have helped me enormously this morning in locating faulty image panes as well as their angles in a huge 75 pane mosaic of the entire Milky Way I have been working on for almost a year. Since I have never been able to get "plate solve and centre" working at this focal length I've had to rely on EQMOD's 3 point alignment method which is often inaccurate. I also look forward to using NINA and Greenswamp server in the future and utilise its manual rotation feature. I will try using your settings on my old 135mm focal length images which I also do a lot of mosaics with but also never had much luck in plate solving in the past.
Thank you so much for your help. It was very confusing knowing which database to use.