IC1805 in exuberant colors + complete workflow

[Herbert West]

Bad weather and Covid-19 isolation forced me to turn to reprocessing my old data. In addition, I had to give my wife some sort of Valentine's Day gift while we're all unable to leave home. Enter IC1805. Any day when I can explain spending hours upon hours on my hobby as doing it for my lovely wife is a good day. Even better, she was very happy with the "gift" 😉

Now, all the colors... What I aimed for is diversity, with red, orange, yellow, green, blue, magenta-ish and teal all distinguishable. Intensity and saturation are just the byproducts. This is classic SHO combination but with H and O LinearFitted to S and preserved green. Tweaked subsequently with Curves and lots and lots of masks, primarily made from O and S stacks and range masks.

I'm not too happy with RGB stars... Stars are the most difficult thing to get right for me.

Gear:

• EQ6PRO
• TS65Q
• ASI1600MMP
• ASIAir V1
• EAF
• EFW
• Filters: Baader Ha, OIII, SII, Astronomik DeepSky RGB

Subs:

• Ha: 128x300s
• OIII: 136x300s
• SII: 118x300s
• RGB: 50x30s each
• Darks, Flats, Dark Flats.
• November 2021, rural Poland, Bortle 3-4

Workflow (99,9% Pixinsight):

Preprocessing: dark and flat calibration using the WBPP script; NormalizeScaleGradient; ImageIntegration.

 

A. Linear processing S, H, O for RGB channels.

A.1. Mure Denoise on S, O (variance 0.3, cycles 30) and H (variance 0.2, cycles 30)

A.2. DynamicCrop on S, H, O

A.3. DynamicBackgroundExtraction

A.4. TGVDenoise with mask on O, S- medium intensity, and H- low intensity.

A.5. MultiscaleMedianTransform denoise with strong mask on O, S and H- with even brighter mask.

A.6. LinearFit H and O to S.

 

B. Luminance layer processing (Ha stack)

B.1. Copy Ha stack from A.3 stage.

B.2. Very mild TGVDenoise with mask.

B.3. Very mild MultiscaleMedianTransform deonoise on wavelt layers 1, 2, 3, 4 with strong mask.

B.4. Deconvolution round 1 with custom global dark/bright settings, local support-  30 iterations.

B.5. StarXterminator to remove stars and create stars only image.

B.6. Deconvolution round 2 with custom global dark and global bright settings, no local support, increased wavelet regularization, 2 iterations.

B.7. Subtle MultiscaleLinearTransform sharpening on wavelet layer 3 only.

B.8. Very mild MultiscaleMedianTransform denoise with very bright, low contrast mask on wavelet layers 2, 3, 4 to smooth out some middle-low frequency noise from deconvolution and MMT.

B.9. Pixelmath to add stars back.

B.10. HistogramTransformation.

B.11. HDRMultiscaleTransformation (EZHDR script, good enough).

B.12. Starnet2 to remove stars and create stars only image.

 

C. Channel combination

C.1. LRGBCombination

C.2. Invert image.

C.3. SCNR to remove magenta cast arround stars.

C.4. Invert image.

C.5. Starnet2 to remove stars.

C.6. HistogramTransformation.

C.7. CurvesTransformation- a lot of it, with a bit of SCNR, with range masks and masks made from O, S and LUM stacks, modified with Curves and HistogramTransformation, with wavelet layers 1, 2, 3 removed with MMT.

C.8. LRGBCombination to add luminance.

C.9. Again- lots of CurvesTransformation.

C.10. A little bit of LocalHistogramEqualization with very low contrast and low amount (scale 32 & 64).

C.11. DarkStructuresEnhance, low amount.

C.12. Very delicate ACDNR with bright, low contrast range mask to protect the background on chrominance and luminance.

C.13. Export as tiff.

C.14. Topaz Denoise with the following settings:

I masked Melotte15 and other high-snr, bright, high-contrast areas to avoid sharpening, because even with “Enhance Sharpness” set to 1% the results there are totally unacceptable! However, Enhance Sharpness set to just 1% preserves all those very dark fluffy bits in dark areas. Whereas, when set to 0% they get extremely blurred. Medium brightness areas seem to be untouched by sharpening so there should be no artificial details introduced by Topaz here.

C.15. Import to Pix.

C.16. A little bit of Curves.

C.17. Pixelmath to add stars.

C.18. Clone image with stars, EZStar Reduction on clone (A.Block Method).

C.19. Mix original image with stars with C.18 with Pixelmath: .7*reduced_stars+.3*org_stars

 

D. Stars

D.1. Pixelmath to combine R, G and B stacks.

D.2. PhotometricColorCalibration with background neutralization- but it didn’t go well.

D.3. HistogramTransformation.

D.4. Starnet2 to extract stars only.

D.5. SCNR to remove green (yes, the stars were greenish).

D.6. Slight Convolution.

 

 

Feel free to ask whatever questions you want answered and post any and all critique. Thanks in advance!

Edited February 15, 2022 by Herbert West

[glafnazur]

Super image 👍