Now this image has so many things going with it, hence the title: testing the limits. So what is so "special" about it? Well, here some highlights: 

1. Another super rare clear (half) night in the middle of August, I can truly say: unheard of in Florida. I had to make it short since I knew that clouds and rain are closing in, hence just 3 hours of NB data.

2. Had to get some RGB data for the stars, got them 3 days ago. I did not want to spend too much time on those as well, resulting I think in a minimum of integration time for a 2022 target.

3. I wanted to have a combined image of the Bubble and the Lobster, but the rest of the image was really not looking nice, hence I cropped the Bubble dramatically. Certainly a little "byproduct" of < 4 hours.

4. I processed this one from scratch around 10 times, all with different channel contributions, mainly stretch changes. This I think is a record for me. I managed to combine 4 of them into one.

5. In order to lower the Bubble in brightness I merged specific areas only and added them carefully.

6. The first addition of the RGB stars left a LOT of stars in the dark, just did not look right. I know a lot of people try to reduce the stars to an almost invisible amount, gives more attention to the nebula, not for me this time.

7. In order to get all of the stars that I could get with the RGB channels I stretched the image a lot, and then masked out the blown out stars, with some new tricky masks that I created. Turned out pretty well, I will start doing this more often to make the process gradually better.

Process description: Take a low stretched star image and blend it into the starless image, then stretch the stars only image hard. Generate a star mask leaving out the small ones in the settings, then stretch it accordingly so that most of the very bright stars are only visible, then apply a convolution on this star mask (25-30), set the mask on the star image and apply 0 light to it (meaning anti stretch them to nothing, not sure what word there is for that 😊😊). That will "kill" all the bright stars, leaving just the dimmer ones, which you blend into the previous image with the low stretched stars.

8. I resampled the image by 4 with Topaz Gigapixel, otherwise the image would have been just too small. Brought out some details as well. Stars as always a challenge when working with Topaz, hence I did some special tricks there too. When plate solving the final image I had to pass 3640mm as focal length due to the 4x ... now that was special, gave me a smile on my face 😊. Look also the pixel scale (0.213 arcsec/pixel). A field radius of 0.263 degrees, that is almost Gary Imm land 😊.

So here it is, less than 4 hours (for which I am more than happy to get in August), using almost every processing knowledge that I have to get a decent image out of it. Could it be better: oh hell yes, especially with more data, but also ... I believe ... a much larger focal length would be great to have for this target.
* This image was not submitted for IOTD consideration