# Unable to get ImageSolver to work

The output is from the Process Console - the final output from ImageSolver.
ImageSolver needs two inputs: an estimate of the image centre, and an estimate of the image scale (how many arc-seconds per pixel).
The approximate position of M65 is more than precise enough as a start point for ImageSolver.
The image scale is usually specified as the approximate focal length and the approximate pixel size. If you have an Astrometry.net solution it will include the image scale, which you can enter into the ImageSolver interface.
Since I didn't have the image scale, the pixel size or the focal length I estimated the scale from the image.
My planetarium software (Cartes du Ciel) tells me that the angular distance between M65 and 73 Leonis (the brightest star in the image) is about 46' 37.1". Reading the positions out with the cursor in PixInsight, and applying Pythagoras, I find the distance between them in the image is 1076 pixels, giving me about 2.6"/pix.

... of course, usually you will know your approximate focal length, and you can look up the pixel size for your camera, so you don't need any calculations.

I guess that's why I've never seen it! It hasn't solved.

Will keep trying. Thanks!

Obviously I am an idiot.

I placed the coordinates with intial seed:
Center coordinates: RA = 11 18 55.908, Dec = +13 05 32.30
Resolution: 2.600 as/px
(using M65 search for start position,and a manually estimated resolution)

and it would not solve.

I assume: Script/Image Analyasis/Image Solver....correct?

I check active window, topocentric and I use Gaia DR3 (XPSD) with Limit Magnitude and Distortion Correction?

What else is there??

Sorry. Appreciate the help here.

Marc

Screen shot of my values in Plate Solve.

#### Attachments

• PlateSolve Table M65.JPG
126.1 KB · Views: 34
This is my IS setup immediately befor running:

I don't usually specify "topocentric" since it makes no discernable difference when solving deep space (non-solar system) images in ICRS.
Since I don't know when the image was taken I set the time to "now", but again it will not make any difference to solving.

Obviously I am an idiot.

I placed the coordinates with intial seed:
Center coordinates: RA = 11 18 55.908, Dec = +13 05 32.30
Resolution: 2.600 as/px
(using M65 search for start position,and a manually estimated resolution)

and it would not solve.

I assume: Script/Image Analyasis/Image Solver....correct?

I check active window, topocentric and I use Gaia DR3 (XPSD) with Limit Magnitude and Distortion Correction?

What else is there??

Sorry. Appreciate the help here.

Marc
This works for me. No topocentric because I don't know your lat/long. Didn't change anything from the defaults. Just put in a plate center and pixel scale.

I don't think it will make any difference, but are you really in the Takla Makan desert? (the topographic coordinates default to E and N).
But I'll try with those coordinates and see if it makes any difference.

Update: It solves fine for me in the Takla Makan desert!

Last edited:
What exact failure are you getting? Could you screenshot the process console after your failure?

I am NOT in China! lol.

Hmmm...did not make a difference.

PI Core Version 1.8.9-1 released 5/18/22, build 1556.

I will send a screen shot of the error.

Is this sufficient?

#### Attachments

• PlateSolveError.JPG
198.4 KB · Views: 31
It says unable to find an initial set of putative star pair matches [?]

This is really puzzling. Just as a check, are you running on exactly the same image as us (i.e. the .tif that you uploaded)?

Yes.

Puzzling, correct. Frustrating too!

I am doing correctly, I assume:

Script/ Image Analysis/ ImageSolver....??

I just plate solved a set of data I took the same evening with the same scope on M44. !!??

Crazy.

I am doing correctly, I assume:

Script/ Image Analysis/ ImageSolver....??
Absolutely! I can't see any problem in any of your screenshots. I am (currently) out of ideas (which is unusual for me )!

the same scope on M44.
That should work really well. Folks don't often take really wide field images of M44, so they miss how huge it is!

Appreciate the help. Will return if I ever figure it out.

Appreciate the help. Will return if I ever figure it out.
Just curious how you obtained this image. As an 8-bit per channel RGB image it's really not very appropriate for any sort of serious analysis. The filename suggests it's been through some kind of PI processing already, but then I'd expect an XISF file in 32-bit floating point, or if saved as a TIFF at least 16-bits per channel. That doesn't explain the weirdness of others being able to solve the image when you can't, but if you have this image in a better format, that would be the one to be working with here.