Two-and-a-half years ago, I captured this image of M31 on my first-ever dark-sky trip. I dragged my wife and my scopes out to Big Bend Ranch State Park next to the national park in southern Texas. Since I was using an ASI294MC with the C11+Hyperstar, I had to shoot this as a four-panel mosaic. When I got home and processed the image, it was a total mess:



At the time, I tried everything I could to make this into something, but it was a no go. I was very disappointed because of all the effort that went into capturing it. A couple of days ago, I decided to dig up this data and see whether there was anything I could do with it. At first, I had no success. The stacks just kept coming out awful--some even worse than the example above. Finally, I decided to try it without the flats and flat darks I'd shot for it at the time. That resulted in a big improvement. And then I also realized that the darks I'd used were taken at the wrong gain. I no longer have the ASI294MC I used for this, so taking new darks wasn't an option. I was worried about the amp glow from that camera, but I stacked all the panels again without any calibration frames whatsoever. WBPP still didn't produce a good result. But APP actually produced something usable. There were some challenging gradients from the amp glow, bit there was minimal vignetting and no apparent dust spots to worry about. 

Even then, APP didn't stitch the mosaic very well. It's normalization function produced horrible banding along the seams. But PixInsight stitched it just fine and gave me enough of a starting point that I was able to produce the final image above. I don't expect this image to set the world on fire, but it's very gratifying to learn that all that effort in November of 2020 to drive out to that park and brave the heat, cold, and the 30 miles of horrible road to get to our campsite resulted in something. Note that I supplemented this image with a bit of Ha data I shot a year later on this target.