Celestial hemisphere:  Northern  ·  Constellation: Cygnus (Cyg)  ·  Contains:  34 Cyg)  ·  34 P Cyg  ·  Crescent Nebula  ·  HD191024  ·  HD191045  ·  HD191176  ·  HD191225  ·  HD191257  ·  HD191290  ·  HD191395  ·  HD191396  ·  HD191397  ·  HD191424  ·  HD191472  ·  HD191703  ·  HD192003  ·  HD192020  ·  HD192041  ·  HD192078  ·  HD192079  ·  HD192102  ·  HD192123  ·  HD192163  ·  HD192182  ·  HD192303  ·  HD192361  ·  HD192422  ·  HD192443  ·  HD192444  ·  HD192536  ·  And 443 more.
Crescent and Soap Bubble Widefield, Kevin Morefield
Powered byPixInsight

Crescent and Soap Bubble Widefield

Crescent and Soap Bubble Widefield, Kevin Morefield
Powered byPixInsight

Crescent and Soap Bubble Widefield

Equipment

Loading...

Acquisition details

Loading...

Description

The Cygnus region is rich with targets and that's a good thing since it's about the only area I can image between the trees in my city backyard!  I was pleased to see the Soap Bubble come out clearly in the Ha and OIII data desipite the city lights.  Maybe not surprisingly, there is not a trace of it in the SII data.  

There is a little "Easter egg" planetary nebula in there.  The advanced plate solve annotation identified it as PK 076 +01.1.  You'll need to go to the full resolution view and it is about 1/3 of the way in from the left, just below the centerline.  It looks much like M57 and seems to have a little glow on either side. 

Another surprise was that Messier 29 is in this FOV!  Never heard of M29?  Well me neither.  It is a very small and not very interesting open cluster in the lower left.  The Astrobin annotations are great learning tools aren't they?

This SHO image used a dynamic combintion as described on The Coldest Nights blog.  https://thecoldestnights.com/2020/06/pixinsight-dynamic-narrowband-combinations-with-pixelmath/  In short, the OIII data is used at a pixel level to decide how much SII and Ha is used in the Red and Green channels, while the OIII data is used 100% for the Blue channel.  This not only produces a nicely balanced color palette, it even gives you nearly normal looking star colors.  So you don't have to desaturate magenta stars or sacrifice SNR by eliminating 95% of the Ha data with SCNR.  In this case I shot RGB star data, but the improvement was not dramatic over the stars that came out of the dynamic SHO combination.  

My basic steps for the dynamic combination were these:

1) Use StarXterminator to produce starless versions of the SII, Ha and OIII masters
2) Create a Pseudo-Luminance of Ha*.4 + OIII*.4 + SII*.2
3) Linear fit the Ha and the OIII to the SII data
4) Create a "factor" image by converting the OIII to non-linear with STF and Histogram Transform, then denoise it with NoiseXterminator, and finally add a little contrast with curves
5) Run the Pixelmath combination just as described on The Coldest Nights.  I tried some alternate mixes but the best turned out to be what is suggested on the blog post.

I used a Pseudo-Luminance so that I could optimize SNR by using more of the Ha and OIII data.  This also meant I could control the Luminance content better than if I just relied on whatever mix produced the best color.  The best color data and the best luminance data are not the same thing!  After simple STF and Histogram Transform stretches, I combined the Pseudo-Luminance and Dynamic color data using LRGB combination.  I did use chrominance noise reduction and a slight bump in saturation during the LRGB combine process.  

I then combined the RGB star data, color balanced with PCC and ran star exterminator on an RGB star clone.  I subtracted the Starless RGB from the RGB with stars data and that gave me an RGB stars only image with a zero ADU background.  This stars only data was stretched and saved as a TIFF for later use.

Moving to Photoshop, I opened the starless Dynamic SHO image and ran NoiseXterminator on it.  This was a fairly light application of noise reduction, I think around 50 noise reduction and 20 detail (but don't quote me on that).  I then used the Camera Raw - Color Mixer tool to push the Yellows and Oranges toward the Red.  My goal was to have as many sublte hues as possible, each represently some diffferent presence of S, H, and O.  

A High Pass filter at around 100 pixel scale was applied with a soft light blending layer to bring more contrast.   

The RGB star data was added with a Screen blending mode.  In addition, the star data was used as a mask.  This all functioned to both insert the RGB stars and also reduce the size and strength of the stars.  Since the RGB stars were a separate layer, I could continue to process the nebula without impacting the stars.  I was also able to change the brightness of the stars to my taste by adjusting only the star layer.

Various other finishing touches were applied, but these are the fundemental steps I used to get to this final image.  Thanks for following along!

Comments