Lots of experiments behind the scenes of this image. This image is my use of the Chroma LoGlow filter to try to control the background glow of Phoenix to my NNW to SSW. What a great improvement! I highly reccomend.

Comments for imagers using the Gsense 4040 sensor (or similar).
My other comments are related to experiments crafted to reduce/eliminate residual fixed pattern noise with the Gsense 4040 sensor (or similar in the case of the Aluma AC4040) as well as determining if HighStack Pro (on camera stacking) is useful for my purposes when imaging with the Diffraction Limited Aluma AC4040 camera.

1) The promise of the sensor is that it is very sensitive and has low absolute read noise, even at fairly high cooler temperatures (0 to -10C). This comes at a cost of a very strong fixed quadrant pattern that is very hard to control with darks and flats. The chip also has ghosting issues. I measure a half life of about 3 minutes for ghosts of bright stars. Also, if you take a flat to 50% saturation, and then take darks, yo get a residual flat on the dark with a 3 minute half life. I suspect this is also an issue with sky flats when saturating the chip creates a ghost of the flat which takes several minutes to attenuate, ie, adds into the next flat. There also appears to be variable read noise that causes the chip to read differently when exposed to light than when the chip is fully dark. This causes dark subtraction to under-represent the pattern noise present in the light frames, leaving faint patterns in dark corrected frames and dark corrected flats. Whether the ghosting or the variable read noise is predominant, the result is also that the chip pattern gets imposed on the flats and therefore appears in the calibrated light frames. My experiments show that the higher the median value of the flat, the deeper the pattern noise remaining in a calibrated light.

2) The Diffraction Limited HighStackPro algorithm has great promise, based on the fact that on camera stacking can increase the dynamic range of the sensor (12 bits native or about 3700 ADU in High readout mode). HighstackPro divides the total frame into multiple shorter sub-frame reads, stacking them electronically on camera. For instance the Ha frames in this image were 300s exposures which the camera divided into 10 30s sub-frames. Signal to noise increased up by the sqrt(10) and the dynamic range increased from 12 to 15 bits. That said, even with dithering of +-60 pixels, the fixed pattern remained in the calibrated and final stacking of HighStackPro frames. Side by side, stacks calibrated with identical flats, HighStackPro vs High mode acquisition (with darks taken in corresponding modes) revealed the HighStackPro mode "deepens" the pattern noise and leaves more residual behind. Note that for this image I added HighStackPro Ha subs to the galaxy through a mask and the effect of pattern noise in Ha was eliminated in processing. If the framing were different, I would need to throw away the HighStackPro Ha and reacquire in High mode.

My Practical IMAGING Conclusions for the Diffraction Limited Aluma AC4040
A) Work the cooler as lightly as possible. The chip has great noise performance and the cooler introduces some banding. Keep that power in to 30-45% range.

B) Take flats at no more than 10% saturation of the chip. I used 350ADU for the flats used in this image. At 800 ADU, the flats reintroduced the pattern noise. I realize this low ADU technique produces noisier flats but the pattern noise is reduced which is a better outcome. Better to accumulate more flats at lower ADU. This image has 9 flats for each of East and West rotator positions.

C) Dither to as large an amplitude as you can. The dithering amplitude was over 60 pixels for this image. This dither DID NOT fox pattern noise for HighStackPro frames, but DID fix it for High mode acquisitions (along with A and B above). The dithering also helps eliminate the residual ghosts around brighter stellar features.

D) My default imaging will be done in High mode and not HighStack Pro for the foreseeable future. I idon'thave a way to eliminate the pattern noise that is somehow amplified by on camera stacking in the Aluma AC4040. If I need higher dynamic range, I'll take exposures at different time and combine using HDR composition.

Message me if you want to know more.