This image represents an improvement in my acquisition process for the Aluma AC4040 camera with GSENSE4040 sensor.  I had noticed pattern noise creeping back into my Lum images, but not my RGB or Narrowband. The average brightness of Lum frames was almost double that of the RGB frames in recent images and I suspected that was an issue with this sensor ( the GSENSE4040 has both RBI and non-linear read characteristics). Consulting with @Howard Trottier (thanks a bunch Howard!), who runs a FLI camera with a similar chip, he suggested I acquire my Lum flats to roughly the same average intensity as the Lum frames. When I tried this, voila, the pattern noise in the integrated Lum images was eliminated. So once again the GSENSE4040 is very sensitive, but behaves in very different and nonlinear ways than other CMOS chips.

I suspect there is another breakthrough to be made in the HighStackPro mode for the Aluma AC4040 camera, where on camera stacking allows for sixteen 12 bit images to be added together on camera before the final readout of the image. This allows 16bit dynamic range instead of 12 bits for the single frame mode and increases overall S/N ratio. Unfortunately, I have not been able to eliminate pattern noise in HiStackPro but now I suspect the average intensity of a subframe fed to the on-camera stack needs to be used as a target for the average intensity of flat frames. Howard's work has shown that this intensity matching of flats is the only way to get a flat that matches the intensity related readout of the data on the chip, allowing for this associated pattern noise to be calibrated out. I hope to get to those experiments in May to see if I can open up the ability to use HighStackPro without pattern noise.

And after the above discussion, I'll just say that M88 was next in line for my winter-spring tour of the Messier galaxies between the clouds, rain, wind, and moonlit skies of the local 2023 Arizona climate. Processing life is easier when patterns are not present in the background.