Mosaics in APP - How I did the recent M45 set

[Scotty Bishop]

I am surprised that I haven't been seeing more of the M45 set done as a mosaic, but then I heard that some were having trouble with it. Here I'll tell you how I was able to do it in APP.

First, you need to stack each panel individually, and on the registration tab (tab 4) make sure that you have "same camera and optics checked" just like you would with a regular integration. Do a new run each time rather than throwing each panel in as a session. Once you are done with that and ready to do the mosaic you will load them all as lights using only one session.

Now to do the mosaic you will need to make some adjustments. Go to tab 3 and set your stars target to 2500, go to tab 4 and set scale stop to 15 on the top end, check "use dynamic distortion correction" uncheck "same camera and optics" and change registration mode to "mosaic." On tab 5 Normalize set mode to "advanced."

 

Once you have done all this, go to the integrate tab and hit integrate. I like to set MBB to 15% on a mosaic and LNC to 2nd degree 3 iterations, but you can do what you want here. A word of warning, this might take a while. Once done use the batch rotate/resize to get it set to where you are able to crop it how you want to, then batch crop to cut off the edges with any stacking artifacts, then the remove light pollution tool to clear any gradients, then put it in whatever program you like to finish it up or even finish up in APP if you want to. I use Photoshop for my finish up work. Don't forget to post your Telescope Live sourced mosaics in the gallery! Here are my 2 latest posted in the gallery along with screenshots of the big LMC mosaic in SHO and HOO that I have only done the initial processing on and will post in the gallery when they are done.

https://telescope.live/gallery/m45-mosaic   https://telescope.live/gallery/horsehead-area-mosaic

 

[Alexander Curry]

Thanks for the guide! This seems a lot more straightforward that doing mosaics in PI 😂

[Jarmo Ruuth]

That was a nice data set! I have heard good things about APP when creating mosaics. It certainly looks easy when compared to PixInsight.

I usually create mosaics in Photoshop or Microsoft Image Composite Editor. This time both failed to create a mosaic so I had to do it manually in Photoshop. Doing it properly inside PixInsight would be a better way but it is so complex that I rarely bother to do it.
 

[Philip F.]

I really need to try APP. 

I suspect that one of the main reasons that people struggled with the M45 mosaic data set is that a couple of the panels will not load into WBCC.  It took me a while to figure it out.  Basically, the coordinates in the FITS header end in 60 and pixinsight expects it to be <60.  I had to individually edit the FITS headers on many of the individual lights to get everything loaded up into pixinsight.  I wish there was a way to add notes to the individual observations to let others know that this is the issue with these files and pixinsight. The next challenge with the M45 dataset was that I went looking for the photometricmosaic tool and discovered that it had disappeared from pixinsight.  Time to learn a new way to make a mosaic!

You mosaics look good.  I might have to try out the horsehead mosaic next.

Cheers,

 

Philip

[ERNESTO GUIDO]

Hi Philip, 

We discussed this WBPP issue with coordinates in the fits header at the end of this thread (and few months ago also on the facebook group). You can easily edit all the fits in a few seconds with the software cited in that post:

 

 

Shortly, I'll start a new thread to discuss this issue with all the info so that all users will be aware of the solution when they encounter it. 

 

Clear Skies,

Ernesto

[Scotty Bishop]

Here's another one that I have to really hand it to APP on, the LMC mosaic. I used all 8 datasets plus the NGC 1763 dataset for a total of 20 hours per channel or 60 hours total integration time. Even though this looks HOO, it is really HOOS because I used the Sii to create false color RGB stars and clusters in many places.

[Alexander Curry]

Nice job! Is the 1763 dataset from a different scope? If so how did you find it handled the differing pixel scales?

[Scotty Bishop]

Nope, it is from AUS-2-CMOS too, so it was a perfect fit on the set. Here are the observations: https://app.telescope.live/click-grab/all?telescope=63&target=NGC 1763