Telescope Live Mosaics

Rex Alldredge

I have spent some time processing two mosaic from Telescope Live:  A six panel mosaic of the Large Magellanic Cloud and a nine panel mosaic of M45, the Pleaides.  I have some observations, at least on my computer.

1.  No mosaic produce by PhotometricMosaic has a valid astrometric solution.  I don't see why this should be so, since things are being assembled by coordinates.  Perhaps there is a format issue between upgrades.

2.  I have successfully obtained an astrometric solution for 1x2 mosaics, 2x2 mosaics, 2x3 mosaics, I have not been able to obtain a valid solution for 1x3 mosaics.  I have examined the fits headers and don't see why this shouldn't work.  This means all mosaics with an odd number of panels cannot be processed.  Specificaly my M45 Mosaic of the Pleaides.  The best I could do was an 8 panel with a chunk missing from one corner.

3.  My larger mosaics seem to have holes in them.  These are small utterly black areas typically near the joins (but not at).  My 3x2 mosaic had one, the "8" panel M45 mosaic had two.  I have noticed that PhotometricMosaic gives a truncation warning from time to time.  I don't know what this is or what to do about it.  Or if it is even related.

Since the areas are small, I've tried to Clonestamp them out of existence with mixed results.  I think Pixinsight could really benefit with something like the Photoshop "Healing Brush".

John Murphy

Hi Rex,

(1) PhotometricMosaic should preserve the header information when it creates a mosaic. However, I think you will find that the MosaicByCoordinate script does not populate the registered files that it creates with the astrometric solution.

(2) This sounds like a bug in the MosaicByCoordinate script. I believe this script has been updated for the latest version of PixInsight. This will have been necessary because they changed how the astrometric solution was stored.

(3) If this is related to a truncation warning, this would mean that when it normalized one panel to another, there was insufficient head room when subtracting the gradient. If this is the case, you can solve it by using PixelMath to add a pedestal to all the input images (you have to use PixelMath because it is vital that black (zero) areas remain at zero), or choose a different reference image that has a higher background level.

Regards, John Murphy (PhotometricMosaic and NormalizeScaleGradient author)

https://astroprocessing.com/