This is an experimental picture. I would like to try how much I can get into Rigel's bright halo and reveal the nebulas (if any) around Rigel. To do so, I first picked my new AP92 Stowaway, which has pristine optics. I probably can't do it easily with my other refractor (15-year-old, lots of dusts and whatever other fine things on the lens elements).

The bright halo caused by Rigel contains several components. First is scattering caused by atmosphere and whatever imperfections on the lens elements. The halo caused by such scattering should be axisymmetric. Then there are multiple reflections between the sensor surface and flattener lens elements, combined with diffraction effects from the micro-lens arrays on the sensor. This leads to faint halos and spikes aligned with the sensor. This is not axisymmetric, and only symmetric about X and Y axes. To model such a halo, I first removed the stars using StarNet++, and then tried to model the halo according to its symmetry. I only removed 80% of the halo because a 100% removal does not look natural. Also, as can be seen from the picture, there are reflection nebulas to the south-east (lower-left) of Rigel. These nebulas interfered with the modeling of the symmetric component of the halo, so a 100% removal leads to over-subtraction in the north-west (upper-right) side. To maintain the correct brightness and color of objects hidden by the halo, all these were done during the linear stage.

Rev.B shows a comparison of simply stretched images before and after the halo model subtraction. It can be seen that, without such subtraction, it is not possible to see the blue reflection nebulas around Rigel.

After the halo subtraction and some processing, I added a deep stack from 20-second short exposures, to fill in the saturated pixels in the 5-minute exposures.

There are still many structures around Rigel, such as the radial spikes and rings. The radial spikes are there because they are not part of the model. The rings should be part of the axisymmetric model. They are still there after the model subtraction likely because I only did an 80% subtraction. So there is still room for improvement. Nevertheless, by subtracting the axisymmetric and X/Y-symmetric components, we can see much more clearly what are typically behind the bright halos of Rigel. Before this, I was not aware about the blue reflection nebulas. Maybe one day I will do the same to Betelgeuse. We will see.