Hi all!
I have quite small FOV with my new 200mm and even smaller with my old 300mm and sometimes I'd like to image a greater area without a lot of hassle calculating coordinates, so some kind of automated mosaic imaging help would be nice.
Br
/Markku

It just so happens that this "wishlist" coincides with something I really need! I was capturing a picture of the heart nebula SH2 190, but due to my limited FOV, I can only get part of it.

Therefore, it seems I'd have to work on an easy way to create <em>job requests</em> that you can feed to the scheduler to take batches of images to enable you later to create a mosaic. I didn't even begin the thought process about implementing this, so any ideas are welcome now!

If you crack this problem, it certainly will be an awesome feature.

There's multiple aspects to the problem. One is how to define the overall mosaic area you want. I can think of at least 4 ways that could be used:

* Use the kstars view to mark the area.
* Use an existing image and plate solve the are details
* Specify "top-left", "bottom-right" coordinates for the bounding box.
* Select an object, e.g. M42, etc and use the object's data.

From the area, you could then calculate the theoretical FoV images required, with whatever overlap % is requested.

One issue I can see is where you may need to adjust the FoV to get guide stars, especially if you need to rotate the camera position.

Then for actual acquisition, you may want to gather by filter per FoV or the other way, by FoV per filter.

Nice ideas! I think some of those could be incorporated. The user may initial Ctrl+F to find the object, and if selected, then the "Top-left" and "Bottom-right" coordinates are automatically selected for 1x FOV with the object being in the center, then the user can edit those fields to change the coordinates, but at the _same_ time, has an option to drag and expand a box on KStars map.

Perhaps this can be aided by another feature I had in mind for KStars, FITS overlay. But this is altogether another big project even though it would be extremely awesome to have FITS overlay from online sources. Speaking of which, anyone knows of any reliable sources for data that provide such FITS file (with WCS coordinates) for most of the sky at decent resolutions? And can be queried online or downloaded?

Now this I'm not sure about. Do folks usually process per-filter super-mosaic first? and then the mosaic-per-filter to produce a color image? Or it goes the other way around? i.e. Produce a color image per FOV, then combine each color image into a mosaic? I'm thinking it's probably the first method since you'd probably want to create the mosaic before any processing of the images?? I'm not sure.

I haven't done much with mosaics so far, because I found it difficult to coordinate the necessary multi-night acquisition, plus the odd overlap angles I got from having to adjust the camera to get guide stars in my light polluted skies.

I made my mosaic images from each filter, then combined those together for a final image (and cropped as necessary). I did this because I then had consistent levels within each filter image across the whole mosaic.

However, I can see the other approach would be easier to put the mosaic together assuming you equalise the image levels in each sub frame (not difficult).

With Ekos Scheduler, multi-night acquisition _should_ be quite easy. The Mosaic Tool would generate a set of jobs for the scheduler to execute. Each job has an associated Sequence File to capture the images in whatever filter specified, so these can be used across jobs. You save the jobs to an Ekos Scheduler List file (.esl) and every night you are ready to observe, you just load the file and hit "start" and the scheduler would pick off from where it was stopped the prior night.

Since the FITS Overlay idea might take a while (if it's ever done!), I think we should start with a simpler Mosaic Tool:

1. Select Target: Name, J2000 RA,DEC (already in Ekos Scheduler)
2. Specify Top-Left and Bottom-Right coords. Enable the user to specify the coords from the sky map (like an image program select or crop tool)
3. Overlap %
4. Generate jobs?

Perhaps add some graphics to show the overall target FOV with superimposed user's FOV and how they overlap and fill the mosaic? The image acquisition details are in the Sequence File. Is there anything else that is critical to the mosaic operation?

Mosaics bring in all sorts of issues, not just in the aquisition phase, but in processing too. It's trivially easy if you have a perfectly flat, undistorted field, but, that camera / telescope combination doesn't exist. The larger the panel count, the more issues there are in stitching them together. My understanding is, there are some tools in pixinsight that do a great job of stitching together mosaics, and they do take into account field distortions.

Another important detail when planning a larger mosaic, is how much overlap between the sections ? There are extremes on the method, and variations between the extremes. On one end of the spectrum, using a 4 panel example, it is shot as 4 distinct fields, with a minimal amount of overlap between them, just enough to get alignments. On the other end of the spectrum, you can shoot a series with a small bump between each shot, say 10% of the field.

To choose your strategy, you need to first plan out how the whole thing will be stitched together after the fact. Do your tools stitch them all together during the stacking phase, in which case the 'small bumps' strategy will work. Or, do you shoot it as specific panels, then stack each panel, and stitch it together after the panels have been stacked ?

I think the method we will ultimately end up trying here on our first attempt at mosaics, is to go with 50% panel overlap, and shoot individual panels. So in essence, take one centered shot which becomes the 'reference frame', then build up panels where the 4 corners of the reference frame become the centers of the 4 panels to create a 2x2, then possibly a similar spacing to do a 3x3.

I've been following the thread with great interest but haven't got much time because of all work that must be done before xmas.

For some time ago I got a idea about creating a 3D sphere that I could zoom/pan and drag and drop my FITS images to, also get them to correct places in the sphere reading RA and DEC from the image and create a record in SQL database were the images would be stored and get an easy way to make backup and keeping track of the sky I've covered.
New images at same coordinates would increase some kind of "version number/date" to keep track of changes in the sky. (finding comets)

Well, that's a lot to do and I don't expect anyone to do that, but I tried, but failed due to lack of programming skills maybe I try again some day.

btw, that's a really great image Jasem!

I think the Mosaic Module will be a really great tool!!

Thanks for all your efforts!

I created a simple Mosaic tool that generates the required observatory jobs for you depending on your settings. It still needs testing and I'll test it in Ikarus observatory once the weather clears up.

Great!! I'll try that as soon as the weather clears up.
Thanks a lot!!

I just did my first try.. but it didn't end very well.. but maybe I missed something, its my first try after all.. ??
Mount slewed as it should but no images was displayed.
I also noticed a strange behaviour when imaging manually, doing a 100s exposure produced 2 images in the fitsviewer, 1 image 100s and a second image 1s. ?? I've never seen that before

Here is the log..

2015-12-29T17:27:22 Shutdown complete.
2015-12-29T17:27:21 Scheduler complete. Starting shutdown procedure...
2015-12-29T17:27:21 4 jobs aborted
2015-12-29T17:27:20 M45-Part3 observation job terminated due to errors.
2015-12-29T17:27:20 M45-Part3 capture failed!
2015-12-29T17:26:41 M45-Part3 capture is in progress...
2015-12-29T17:26:41 M45-Part3 slew is complete.
2015-12-29T17:26:36 Slewing to M45-Part3 ...
2015-12-29T17:26:34 Found candidate job M45-Part3 (Priority #10).
2015-12-29T17:26:34 M45-Part3 observation job is due to run as soon as possible.
2015-12-29T17:26:33 M45-Part4 observation job terminated due to errors.
2015-12-29T17:26:33 M45-Part4 capture failed!
2015-12-29T17:25:54 M45-Part4 capture is in progress...
2015-12-29T17:25:54 M45-Part4 slew is complete.
2015-12-29T17:25:49 Slewing to M45-Part4 ...
2015-12-29T17:25:47 Found candidate job M45-Part4 (Priority #10).
2015-12-29T17:25:47 M45-Part4 observation job is due to run as soon as possible.
2015-12-29T17:25:47 M45-Part3 observation job is due to run as soon as possible.
2015-12-29T17:25:46 M45-Part1 observation job terminated due to errors.
2015-12-29T17:25:46 M45-Part1 capture failed!
2015-12-29T17:25:07 M45-Part1 capture is in progress...
2015-12-29T17:25:07 M45-Part1 slew is complete.
2015-12-29T17:25:04 Slewing to M45-Part1 ...
2015-12-29T17:25:02 Found candidate job M45-Part1 (Priority #10).
2015-12-29T17:25:02 M45-Part3 observation job is due to run as soon as possible.
2015-12-29T17:25:02 M45-Part4 observation job is due to run as soon as possible.
2015-12-29T17:25:02 M45-Part1 observation job is due to run as soon as possible.
2015-12-29T17:25:01 M45-Part2 observation job terminated due to errors.
2015-12-29T17:25:01 M45-Part2 capture failed!
2015-12-29T17:24:22 M45-Part2 capture is in progress...
2015-12-29T17:24:22 M45-Part2 slew is complete.
2015-12-29T17:24:19 Slewing to M45-Part2 ...
2015-12-29T17:24:13 Found candidate job M45-Part2 (Priority #10).
2015-12-29T17:24:13 M45-Part4 observation job is due to run as soon as possible.
2015-12-29T17:24:13 M45-Part4 observation job is estimated to take 00h 01m 05s to complete.
2015-12-29T17:24:13 M45-Part3 observation job is due to run as soon as possible.
2015-12-29T17:24:13 M45-Part3 observation job is estimated to take 00h 01m 05s to complete.
2015-12-29T17:24:13 M45-Part2 observation job is due to run as soon as possible.
2015-12-29T17:24:13 M45-Part2 observation job is estimated to take 00h 01m 05s to complete.
2015-12-29T17:24:13 M45-Part1 observation job is due to run as soon as possible.
2015-12-29T17:24:13 M45-Part1 observation job is estimated to take 00h 01m 05s to complete.
2015-12-29T17:24:12 Dawn is at 06:12:00, Dusk is at 17:45:00, and current time is 17:23:51
2015-12-29T17:23:33 Mosaic file /home/maxer/FITS/M45_mosaic.esl saved successfully.
2015-12-29T17:23:33 Scheduler list saved to /home/maxer/FITS/M45_mosaic.esl

Scheduler is buggy now, I am currently working on resolving a few issues, will let you know when it is ready for testing again.