M106 and some PixInsight insights

[Annehouw]

I have written a short description based on this image and the physics happening around the core region of M106: https://www.astrobin.com/uk8p8v/

It is for those that are interested in the physics of outer space (as I am). In this forum, I would like to say a bit about the image processing part as that might be of interest to some of you using PixInsight. Note that I also posted this on CN.......

 

The goal

This image was processed with the plan to show the main features of the interesting activity that M106 shows around the core. As such, saturation and color contrast have been turned up quite a bit (masked).

 

Ha addition to RGB galaxy data
The anomalous arms (see Astrobin for the explanation) are fairly easy to see in a Ha image, but getting the delicate gas streams well defined added to an RGB image proved to be very finicky as the contrast between them and the red background signal is not that high. Furthermore, this is an image made with an OSC and a STC-DUO multi-narrowband filter. The Ha signal is only captured by the red bayer filtered pixels, hence diminishing the spatial resolution of the Ha part. I struggled quite a bit and after some experimentation with adding Ha in the non-linear phase using image scale operators, I ended up with a manual subtraction of the red continuum in the linear phase, as described in an accessible way by Eduardo Luca Radice   http://www.arciereceleste.it/tutorial-pixinsight/cat-tutorial-eng/85-enhance-galaxy-ha-eng . I used both a "clean" Ha derived image and a range mask from that "clean_Ha" image to add the clean Ha contribution to the image.

Note that the continuum subtraction method only works with linear data! Using this on non-linear data turns the whole area red.  There is also a CS (continuum subtraction script) script that should do the arithmetic of the Ha fraction to the red signal, but that did not give me an acceptable result.I do not know why.

 

Star Halo Reduction

My telescope/filter/camera combination gives me prominent halo’s. I used the method described by Nuno Pina Cabral to reduce the worst of them.  You can find the method here.:  http://nunopinacphotos.blogspot.com/2016/03/how-to-remove-star-halos-using-wavelets.html
Note that I used HDRMultiscaleTransform instead of ATrousWaveletTransform (or MLT/MMT for that matter) as it gave me somewhat cleaner results. The beauty of the method is that it tackles all halo’s in one go.

 

EZ Processing
This was the first time I also used the EZ_suite noise reduction and deconvolution scripts. Works brilliantly (although I did have some PI crashes, but I had been warned….). What a nice concept that is.

 

I hope that this description adds something to someone's PI journey.

 

 

Processing this image took nearly as long as the total exposure time as I had to find my way into the methods above and because I made a lot of  image versions varying the amount of Ha contribution before settling on this one.
Equipment and acquisition details,, together with a larger version of the image and the physics explanation on Astrobin.

 

Edited November 1, 2020 by Annehouw

[Laurin Dave]

Nicely done and good write up..  I see from Astrobin that you have Photoshop and wondered whether you have tried adding the Ha to the red channel using blend mode lighten ? I have Pixinsight but find the PS method somewhat easier and more controllable ..

Dave

[Annehouw]

Hi Dave,

Yes I did try that. It was my first go at this and it would work well once you have isolated the Ha excess signal. Without doing that...since the core of the galaxy is quite bright in Ha, just adding the Ha image to the RGB image resulted in a red core area.  So that was a dead-end for me. I guess you could do the continuum subtraction in PS through a set of subtract blend mode operations, but I suspect that this is one area where PI is less complicated than PS 😉 

I do use PS in my workflow as some things are (like you say) just so simple in PS. First of all to clean up masks I make in PI. The clone tool in PI is clumsy. Futhermore, I do like to make image versions having their own strong points in PI and then layering them in PS, play with opacities and mask at will (either using the masks I made in PI or just using PS masking).

[geeklee]

Great write up and image.  Thanks for sharing the links to the methods used - it's always enjoyable learning and trying techniques in PixInsight that may replace/complement existing workflows.