Revision 1/14/2023:
With this revision, I decided to apply the xXT functions to improve emphasize the Starstreams surrounding M51, the more sizable galaxies in this neighborhood of M51 and sharpen and reduce the starfield to help with the first two goals.  The stars were pretty good in the original, so I struggle in which I feel is better.  But I conclude these two are just different.  I actually like the brighter, colorful, and more populated starfield in the original.  But reducing the stars certainly helps in making the galaxies stand out.  The detail is also enhanced in the new version with the BXT being the primary tool.  I bumped this up so that it was even obvious when viewed at full field.  Yet I feel that it is not so much that if you zoom into just M51 the detail holds as true.  Mostly, sharpening here seems to have allowed for better contrast between the bright areas and the dark cloud lanes and filaments.  It should be stated that unlike the original image, I did not split subs into different quality groups, nor use a best luminance image to improve the final result.  All subs were used for the RGB image and processed simply.  I believe the details on this revision are clearer and more accurate than the original.

Oh, and not to forget, NXT did a wonderful job on the noise without destroying delicate structure.  Key to a final image without the noise seen in the original image.

Old Description:
This is probably the most photographed galaxy ever. Normally I would not offer any background on it, but then maybe most people who post M 51 may think the same! So for those who are interested and for friends and family, here goes: M 51a&b describes two spectacular interacting galaxies about 30 million LY distant. The larger grand design spiral is "a" and is said to be 76,000 LY in diameter. I will assume that this does not include the wide-ranging star streams caused by the interacting pair. While certainly smaller than the 100,000 LY diameter Milky way, it surprisingly has only ~10% of the number of stars as the our home galaxy. However, as can be seen, M 51a has abundant dust and it is undergoing a very active period of star formation. One might think that this is due to the interaction with "b", but to my eye, for the proximity of these two, there is relatively little obvious tidal disruption to the spiral arms of "a", while "b" is profoundly impacted. While my image is intended as a wide field view. It can be seen that the very regular order of the spiral arms and dust lanes extends deeply into the core, all the way up to the very tight stellar core. So I find it unlikely that this galactic dance is the cause of the active star formation in "a".

This data was shot with a OSC camera. I had originally collected well over 200, 45 sec subs on this and from that arrived at 97 after scrutinizing. I was also deeply disappointed that upon calibration and stacking, I got a galaxy that was clearly not as sharp as an image that I rather quickly acquired with my 5 inch RASA quite a while ago. I decided to use a new-to-me method of processing, which involved a harsh sub selection based only on star quality, in particular size and shape to improve the resolution. This lead to 55 "super selected" subs which I used only to generate a manufactured Luminance image. It is this L image that I did my deconvolution, etc. I could see that I had indeed improved the image resolution and this lead to my being able to see a bit more structure, improved results with deconvolution and allowed for better resolution of dust structure. I still used the original 96, 45 sec subs as the source of the RGB signal and processed that to the extent necessary to allow for recombination. I will not go into the details, because I know that you narrowband and RGBL imagers already got this down pat. But for me, I am breaking new ground! In any case, I was stunned at how well the process worked using OSC data. In retrospect, I could have been even harsher with my sub selection for L, but really this is a wide field view and should be viewed as such. I believe that tracking issues and seeing (wide aperture) led me down this road. I may well employ this method in the future, though I recognize that this target's very high relative dynamic above its local background likely helps.

On the color side, the image is photometrically color calibrated. Beyond stretching, there was a bit of contrast enhancement (Curves Trans) in the L channel (after recombination) and some saturation to return some color lost during the L contrast enhancement. I did no differential stretching (with mask) to preserve the core region, yet the crepe/lace structure is evident all the way to the tiny core. Looks like a Buckminsterfullerene for those chemists out there! I did some denoise work on the L channel, since I heard that it could be prominent after recombination. Also did stronger denoise on the RGB image since it was my understanding that only color info would be used in the recombination and color structure would not be impactful. This is part of the reason that I used all the subs for the RGB image. There remains noise in the galaxy itself, because galaxies were heavily masked during the channels denoise effort. I was able to process out most of the noise in the galaxy and the peripheral star stream area, but reversed course because I did not like how it impacted the finer details in the dust lanes and in any case, it is not intended to be viewed at 500%. I know that this information is highly redundant for those who do multichannel imaging and processing, but I thought that doing so would allow anyone who has advice or criticism on what I did, or shouldn't have done can know what my approach was. Plus, since this is with OSC data, I thought that might be of interest.

For QHY users: Mode 1, gain 0, offset 11, which is basically unity gain for this camera.