Mosaic Questions

[niteman1946]

Hi Group,

I'm running PI v1.8.101098, and using Harry's mosaic tutorial video,
http://harrysastroshed.com/pixinsight/pixinsight%20video%20html/pixinsighthomeinter.html

The two tile target is NGC7380, using Sequence Generator Pro's very handy Framing and Mosaic Wizard.  Telescope is a Meade 12"LX200 with a Starizona SCT corrector and F/7.11 reduction.  Focal length is 2182mm.  SGPro creates the two tiles (upper and lower) with a 20% overlap.  While the SCT scope and Starizona's corrector create a fairly pleasing edge image (prior to mosaic), there are still some distortion if you look hard enough.

Mosaic Technique:
1.  Dynamic croped very slightly lower edge of upper tile and upper edge of lower tile to eliminate any dithering crud.
2.  Star Align Upper tile to Lower tile using "Frame Adaptation" and "Register/Union-Separate".
3.  Gradient Merge Mosaic using "Overlay", "Feather Radius of 20" and "Shrink Radius of 5".

Here are the screen shots of the results:
1.  Mosaic Merge Full DC
2.  Mosaic Merge Enlarged along Seam DC
3.  Mosaic Merge Enlarged Greatly Along Seam DC

Image #1 looks pretty good.  You can see some of the remnants of the merge seam in Image  #2.  And you can see quite a bit of very small double star-over-star along the seam in Image #3.

Given my equipment and the technique I used, is this about as good as I can expect with the mosaic capability in PI?

Thanks for you thoughts,

Mark

[chris.bailey]

Mark

Try using the Painting>Clone Stamp tool to remove any bright stars on the very edges of overlap, espeically those that are incomplete. This tends to avoid those dark shadow artefacts in the merged image. Are you enabling distortion correction in Star Align with the 2-d surface splines registration model? This can be very effective at avoiding double stars in the seams.

Chris

[niteman1946]

Thanks Chris,

I will give both of them a try.

Mark

[georg.viehoever]

See http://pixinsight.com/forum/index.php?topic=4105.0 for tips on how to fix the artefacts caused  by GMM.  Unfortunately they are typical for GMM with panels that have bright start on the borders.
Georg

[niteman1946]

Thanks George,

Lots of attention to detail, but I will give it a try.
Hopefully afterward, I'll have something decent to post.

Mark

[Juan Conejero]

Mark, enabling distortion correction and selecting the 2-D surface splines registration model are essential for mosaic construction with StarAlignment. See this article for more information on these features. With these options you should be able to fix all geometric distortion problems in your mosaics.

For wide field mosaics with many frames you should consider the MosaicByCoordinates script by Andrés del Pozo, which is based on astrometric solutions and allows you to generate large mosaics with controlled projections and even mosaics without overlapping areas.

[niteman1946]

Thanks Juan,

I never cease to be amazed at the patience of this group.  I recognize that this material has been covered substantially in the past.

But It's just not been under my radar until I started working with SGPro, and found its ease in acquiring the tiles.

Mark

[niteman1946]

Gents,

The distortion correction and 2D surface splines registration did the trick.

Thanks for all of your help.

Mark

[niteman1946]

Regarding typical "linear" processes ... Do you execute them prior to, or following, the mosaic merging process ?

Thanks,
Mark

[georg.viehoever]

Regarding typical "linear" processes ... Do you execute them prior to, or following, the mosaic merging process ?
...

Usually, I do GMM before becoming non-linear. But you can also do it on non-linear images.
Georg