Mosaic Star Alignment Settings

[Graham]

Hi. 

When creating mosaics, is there any settings I can change in Star Matching, Star Detection and Interpolation to help it solve the panels easier? Although I am working in a fairly heavy star field with what I feel is good overlap, it can sometimes take a long time to solve or even fails as it is not able to find good star matches. Sample of Processing Console attached.

Thanks in advance for any insights.

Graham.

[Andres.Pozo]

I think that StarAlignment is detecting too many stars. You could try to increase the sensibility parameter to decrease the number of detected stars. The default value is -1. You could try with 0 or 0.5.
Also, if the images have noise you could set the "Noise reduction" parameter.

[georg.viehoever]

- Is this a wide field? If yes, you may have run into the typical lens/projection distortion problem. Using the Image Solver family of scripts (http://pixinsight.com/doc/scripts/ManualImageSolver/ManualImageSolver.html as starting point) may help.
- sometimes it helps to place previews into the overlap regions, see http://pixinsight.com/doc/tools/StarAlignment/StarAlignment.html#usage_011 .
Georg

[Graham]

Great, thanks guys, I'll look into it.

Yes, it will be a large area of sky. I am using an 8300 chip camera on a Tak TSA120 to create a 16 pane mosaic.

[georg.viehoever]

At 900mm focal length you should not have problems with distortion resulting from projection. Dont know about lens distortion, but I think it is unlikely.
Georg

[Graham]

Hello Georg,

I'm not sure what could be affecting the mosaic process although I have not had time to read up and understand the other suggestions fully as yet. I am getting some time tonight to have a play and another problem has turned up. I seem to get a misalignment in one corner of the image. With 4 panels it was left bottom, with 6 panels it is top right. I have put in an image for you to see. I would appreciate any feedback from you or others.

I am following Steve Allan's tutorial on Gradient Merge Mosaics.

Thanks in advance. 

[georg.viehoever]

Did you play with the following StarAlignment Parameters:
- Registration Model=2D Surface splines. Gives more freedom in adapting star fieds than Projective Transformation
- Distortion Correction
- Be sure to read the StarAlignment docs http://pixinsight.com/doc/tools/StarAlignment/StarAlignment.html
Georg