For anyone who may be interested, this is my end result.
http://www.astral-imaging.com/M82-2010-1600.htmThe processes used are as follows:
I’m not sure why but M82 has always been a tough target for me and I keep coming back trying to get a good image of it. I was imaging M74 getting luminance data the same nights as M82 was still too low at early night and the M74 data was all below 2.5” arc second seeing. Go to M82 which is over Charlottesville for me (North) and I’m lucky if I can get color data below 3” arc seconds. The luminance data is 46 frames mixed of 1200 and 900 second exposures but 13.3 hours of data. The RGB data is all 20 minute frames and a total of 11 each for an additional 11 hours. The RGB was binned 2X2. The process followed in Pix Insight was to align all the luminance and then the RGB data. Saved aligned files and then combined using the average and median methods with linear fit clipping for the pixel rejection method. Saved the created masters in 64 bit IEEE form as set by the combine tool. Used the Screen Transfer Function tool to screen stretch the image and cropped as necessary on the luminance image and same for RGB after combining the RGB using Channel Combination tool. Used the Dynamic Background Extraction tool to remove any gradients in the images, saving each resulting image after each process. On the RGB image I used the Neutralize Background tool to see if it would alter the result of the Dynamic Background tool and if it did only very slightly and not noticeable. Further processing to the luminance and RGB data was doing a histogram stretch using the track and preview functions. Saved the results and then created star masks for the luminance and RGB images saving those as well. Applied the respective mask to each image and then applied both HDRWavelets and Regularized Lucy Richardson deconvolution to each image again saving the results.
At this point I have a 1x1 luminance image and a 2x2 RGB image. The resulting RGB image is a bit weak on color so I create a saturated color image by using the curves tool and setting it for saturation mode. Saved the result. Now I decide to combine the RGB and luminance images so I use the Dynamic Alignment tool and, using the luminance image as the reference image, pick a star to use as the first point. This results in a green boxed X in the RGB image which is no where close to the star picked in the luminance image so you drag the green boxed X to the proper star in the RGB image and pick another star in the luminance image. The green boxed X in the RGB image will still be off so you move to the correct corresponding second star. By the third star you should be matching star for star as picked in the luminance image. If not you misaligned one of the first two stars and should probably start over. After I pick 8-10 stars making sure they all lined up I’ll execute the tool and get a resulting registered and re-sized RGB image ready to be combined with the luminance. Saved the registered RGB image. At this point after the images have been saved I’ll close all images just to get a clean workspace. I’ll re-open the luminance image and the registered RGB image. Now using the LRGB combine tool I unchecked the RGB boxes leaving only the L box checked. Hitting the down arrow to the right of the checked L box opens a window for me to pick the luminance image I’m using and the only image listed is the luminance image I have opened. Pick that image and it is displayed in the luminance selection. Adjust saturation levels slightly, checked the Chrominance Noise Reduction box and drag the half box icon (triangle) in the lower left corner of the LRGB Combine tool onto the RGB registered image. The result is a LRGB image.
Couldn’t be any simpler than that now could it! What’s amazing about this process is how the time flies when getting involved in these images. What I’ve learned from this process is that once you figure out what settings create a good star mask for your setup (telescope/camera) you can use those most often without needing to change anything. There are also several other ways of creating masks for processing the images which have a slightly different effect. What I haven’t figured out yet is how to do my deconvolution while the data is still in the linear state. I’m told this is when it should be done but from my limited experimenting with deconvolution and linear data leaves nothing I could do anything with. I’m sure I’m missing a step here or even possibly using the wrong deconvolution method for my data. I’m told that the way I’m doing this now is more of a sharpening effect than deconvolution.
