Messier 81-82 in Wide-Field

OTA: FSQ106
Camera: FLI PL16083 (CCD of yore)
Observatory: Telescope Live
Date of Capture: May 21
Date of Processing: May 24

Exposures Used:
R: 16 x 600 sec
G: 6 x  "
B: 8 x  "
L: 16 x  "
Total Exposure time: 7.7 hours
Image Width: xxx arc-minutes

Processing Tools:
1.    Commercial: PixInsight, Topaz, Radiant Photo
2.    Pixinsight Addons: NoiseXTerminator, BlurXTerminator, StarXTerminator, GraXpert
3.    My Scripts: NB_Assistant, AC_Restar, Subframe Weighting Tool (Excel w/ J. Hunt), ColorTweaker, StarTweaker

Target Description:
These are very popular objects, and much has been written about them. Messier 82 is a star-burst galaxy with flares of ionized hydrogen being ejected from its disk. This image does not do justice to the details of that process or the details in either galaxy. It is about the dramatic setting they occupy behind the clouds of our own galaxy. If you want to see an excellent amateur version of Messier 82, try this one by Nicola Beltraminelli.

Processing Description:
What motivated me was seeing Timothy Martin's beautiful rendering of this target from a wide-field view, and I tried to track down a similar data set. This one from TelescopeLive covers about the same fov but lacks the critical Ha data, and the composition of the images misses some of the best parts of the foreground nebula. Another handicap to detail and color intensity was just a handful of each G and B subs (and a better number of R) and a "compensating" abundance of L, which can suppress some degree of noise…but not as much as we can suppress with tools such as NoiseXTerminator. So, L subs are not the S/N heroes they once were held to be. But lacking enough RGB subs, I decided to use them (as described above).

Just playing around. First, I tried treating the L in yet another novel fashion: I made an image from the stacked R, G, B, and L using image integration, then linear fit it to the L extracted from the RGB image and introduced this resulting pseudo-L into the RGB image as we usually do. Unfortunately, I forgot to normalize the 4 images before integration…well, it was just an experiment and generally works. Probably not the best thing to do, however. The posted version is better and introduces the Super-L in the linear image after a linear fit.

Alex Woronow