Another basic mosaic question

[cfranks]

Hi,
I'm  generating a 4-frame LRGB mosaic of Omega Centauri using Georg's GradientMergeMosaic and following Steve' excellent video.  Is there a preferred way of getting to the final LRGB image?  Ie. is it better to create the full mosaic for each of the 4 filters and then use the LRGBCombination, followed by DBE etc, or to make each of the 4 frames LRGB and then GMM them to get the final image, followed by DBE etc.  I imagine the first way would give better control of background, colour saturation etc. but I don't  know enough about the whole process to be sure.
Thanks, Charles

[georg.viehoever]

Hi,

when working with GradientMergeMosaic, you should carefully study the discussion on avoiding artefacts in http://pixinsight.com/forum/index.php?topic=4105.0, and if necessary go through the threads discussing this module http://pixinsight.com/forum/index.php?topic=3572.0.

My impression is that you should create the merged image as early as possible in the processing (before histogram stretch, but probably after LRGB combine), to avoid artefacts generated by unavoidable differences in processing the different panels. On the other hand, merginbg fully processed panels are less vulnerable to the star artefacts discussed in detail in http://pixinsight.com/forum/index.php?topic=4105.0, and it requires less memory.

Maybe others that have done mosaics want to chime in?
Georg

[cfranks]

Thanks Georg,
I am using a fairly generous overlap (10%) and only had one of the black artefacts and the cause was obvious and a larger crop easily fixed it.  I get superb (to my eyes) joins with no visible seams just using the GMM standard defaults.  Of course, with this massive star field, I have no faint nebulosity to worry about but was concerned about the possible differences in the background and faint stellar colours across the frames.  I will follow the threads you mentioned and see how I go.
Thanks again,
Charles