Stitching mosaic panels


#1

Just starting with mosaics and wanted to know what was the best software for stitching the panes together. I know that microsoft has image composite, but don’t know if it handles fits files or tiff. PI seems a little complex. Thanks


#2

I never had much success with the standard PI tools for mosaic creation, but I do use it for the mosaic creation. It’s a very tedious process, but I have found David Ault’s tutorial to work the best: http://trappedphotons.com/blog/?p=994

Basically you create a fake star map and then register all the panels to that star map.


#3

I have done 8-9 mosaics captured in SGP and processed in PixInsight. It is time consuming, but the results are really satisfying.

  1. Find the center of your mosaic and create a synthetic star field.

  2. Crop all of your panels 50-100 pixels, i like to name them Panel1, Panel2, Planel3 …

  3. Used the DNALinear under “Scripts–>Utilities” to make all of the panel the same “brightness”.

  4. Register each panel to the Synthetic star field. I save them as Mosaic1, Mosaic2, Mosaic3 …

  5. Finally use the GradientMerge process to put them all together.

Hints: I like to keep a copy of SGP’s mosaic picture in PixInsight as a reference in orienting my panels.
Make sure to use previews when you do your Registration with your synthetic star field. Look at David Ault’s tutorial. He uses Pixel Math to eliminate star the “pinch”. I have found I can us the clone stamp much faster.


#4

Thank you Joel and Robert for the information. Started two panel using the wizard last night and it seems to be working well. Hopefully this will lead to some great pictures.

One thing I am concerned about is the wcs coordinates required in the fits header. Never plate solved using PI.


#5

Hi Gunny

I’ve used PI for mosaic compilation for many years and am happy for you to PM if you need a pointer at any time.

Hope all is well with you.

Barry


#6

Hi Barry,

Thanks for the offer and I’ll take you up on it when I run into problems. I did a quick msice with some of the images and at least I’m in the ball park with the fsq 106. Really looking forward to this new endeavor.

Hope all is well with you and the imaging site in Spain. CS Barry.


#7

Try Astropixelprocessor from Mabula Haverkamp.
Easy to use with outstanding results.

Kind regards Stefan


#8

Thanks Stefan. I’ve heard a lot of good things about APP, but I already have way too many processing programs now.


#9

you guys know that Andres Pozo’s script called MosaicByCoordinates does all the grunt work automatically, right? all you have to do manually is to solve the input images and do the DNALinearFit step, but sometimes even DNALF is not necessary.

the plate solving in PI works pretty well as long as the images have a good starting hint. in order to use MBC you do have to solve each one of your panes but IMO that step is easy and pretty brainless compared to figuring out all the parameters for StarGenerator and then running StarAlignment…


#10

Thanks pfile. Since I’m new to mosaics, is there a reason why msice seems to be frowned upon verses PI? Just curious as I have yet to use PI for mosaics. There seems to be a lot of steps with PI as opposed to msice.

The two panel mosaic that I posted here and on flickr seemed to be pretty good in my eyes, but I could be wrong…Always willing for this old dog to learn new tricks, so c+c is appreciated.


#11

i have never tried msice - i guess one question is if it can handle linear astro-images. if not then i guess you need to stretch each pane individually and carefully enough that the brightness matches up…

rob


www.mainsequencesoftware.com