Why did I acquire so much Ha for so little yield in the final image (there is a little enhancement). Life (ie stuff I had to deal with) was occurring as I acquired this and I was just able to run the observatory each night with no time for targeting new objects. As the moon rose, I just kept taking Ha to see what I could get. For this galaxy...not much. Moral of the story, always have an extra target speed out just in case.

The short story...I finally have cracked the code of calibrating the AC4040 in High StackPro mode. Very low ADU flats are the key. More details in the long story if you are interested. All LRGB frames are 600s High StackPro and the Ha are 600s High Gain mode (no worries about bright objects needing more than 12 bits of dynamic range)

The Long Story. High StackPro mode is one mode the Diffraction Limited AC4040 can read the GSENSE4040 chip. The drawback for the most sensitive and lowest noise High Gain camera read mode is that it is only 12 bits, so very often, bright areas of galaxies and stars are blown out. High StackPro mode overcomes this by reading and stacking up to 16 images on the camera, allowing for shorter on-camera exposures that preserve detail in high intensity regions, while reading out the final camera-stacked image in 16 bits. The only penalty for doing this is read noise goes up by a factor of 4 since 16 reads of the chip are made for each image. In my Bortles 4 sky...that read noise is not the factor that sets the background in an image. It all sounds good until you go to calibrate the High StackPro image and the notorious issues with the GSENSE 4040 patter noise creep in. For regular images acquired in High gain mode this is handled by acquiring flats at the average ADU of the light subs, which is very low ADU compared to conventional wisdom for flats. However, the low ADU flats work, but you usually need to acquire a few more because they are noisy. I was never able to get this strategy to work for High StackPro mode, until I realized that for a 600s sub, the actual image acquired and stacked on the chip was only 37.5 sec long...much shorter than the light frames I was usually acquiring. So, I decided to take even lower ADU flats and voila...pattern noise dropped away on calibration of the images. Now I can safely acquire targets with 16but dynamic range to achieve better galaxy and star cores.