Getting plate-solving status, please wait...

Globular Cluster M3

Contains: M 3, NGC 5272

Technical card

Resolution: 3540x2360

Dates: March 31, 2014April 5, 2014

Frames:
Baader Planetarium UV/IR Cut 2": 4x300" ISO400
Baader Planetarium UV/IR Cut 2": 6x600" ISO400
Baader Planetarium UV/IR Cut 2": 15x180" ISO800
Baader Planetarium UV/IR Cut 2": 6x90" ISO800

Integration: 2.2 hours

Darks: ~15

Flats: ~30

Bias: ~48

Avg. Moon age: 2.74 days

Avg. Moon phase: 14.04%

Astrometry.net job: 658500

RA center: 205.552 degrees

DEC center: 28.366 degrees

Orientation: -85.880 degrees

Field radius: 0.729 degrees

Locations: York River State Park, Williamsburg, Virginia, United States; LaRC, Hampton, VA, United States

Description

M3 is a globular cluster comprised of some 500,000 stars in the constellation Canes Venatici. This image is a product of two separate nights of data acquisition - one night from a light polluted suburb and another night from a somewhat darker site. Four different exposure lengths are combined with the goal to provide detail in the cluster center while making fainter stars visible.

The PixInsight tools I found most impressive in post-processing of this image were the MaskedStretch script and the HDRMultiscaleTransform process. The former stretched the dynamic range of the image while maintaining stars with soft edges whose colors could later be enriched effectively, while the latter compressed the dynamic range of the cluster to reveal details in the bright center.

Revised processing description
The revised version of this image is a substantial improvement over my initial processing. Previously I had integrated all light frames together regardless of ISO or exposure length which did not help to reveal details in the core of the cluster. Further, I over did it with sharpening the stars which left mostly white stars with abruptly changing point spread functions. This time around I used the PixInsight HDRComposition process to combine images of different exposure length and ISO such that pixels from the least exposed image contribute to the cluster core, yielding more detail throughout. I was also much more gentle with sharpening the stars which kept a smoothly varying PSF and allowed me to push the star colors further than in the original version. Finally, I did not crop as tight this time around to keep that nice orange star in the image because it adds a nice color contrast with the cluster.

-------------------------------------------------------------------------------------------
REVISED Processing Workflow (PixInsight)
-------------------------------------------------------------------------------------------

Pre-processing
1. Calibrate light frames using darks, bias, flats
2. Integrate each set of exposures having matching ISO and exposure time, yielding four master light frames (ImageIntegration)

Post-processing

For each of the four master light frames;
1. Initial crop (Dynamic crop)
2. Reduce background gradient and neutralize background (DynamicBackgroundExtraction)
3. Reduce background gradients more (DynamicBackgroundExtraction; normalize)
4. Neutralize background (BackgroundNeutralization)
5. Set white balance (ColorCalibration; use entire image since it is all stars)

Single HDR image processing;
6. Combine four master light frames into a single light frame, ensuring that pixels from the shortest exposure are used for the galaxy core (HDRComposition; binarizing threshold 0.05)
7. Non-linear stretch (MaskedStretch script; target median background 0.20)
8. Reduce background brightness; increase contrast (CurvesTransformation)
9. Reduce background luminance noise (ACDNR to luminance with luminance mask)
10. Reduce background chrominance noise (ACDNR to chrominance with luminance mask)
11. Set luminance coefficients to 0.333333 for RGB channels (RGBWorkingSpace).
12. Increase color saturation by extracting lightness and then recombining with RGB image (ChannelExtraction, LRGBCombination; reduce saturation slider to increase saturation)
13. Sharpen stars slightly (MultiScaleMedianTransform)
14. Reduce green (SCNR)
15. Increase contrast again (CurvesTransformation)
16. Increase star color saturation (ColorSaturation with range mask)
17. Lower black point to around 0.1 (HistogramTransformation)
18. Cosmetic correction to a few spurious pixels in some stars that were over saturated (desaturation of the specific color with ColorSaturation using mask to select offending stars)
19. Realign histogram peaks in the RGB channels since they slid slightly (HistogramTransformation)
20. Flip and rotate to put the orange star in the lower right (FastTransformation)
21. Final Crop (DynamicCrop)
22. Set ICC profile to sRGB for web publishing (ICCProfileTransformation).

-------------------------------------------------------------------------------------------
ORIGINAL Processing Workflow (PixInsight)
-------------------------------------------------------------------------------------------

Pre-processing
1. Calibrated light frames for all exposures averaged together; weighted by average signal strength (ImageIntegration)

Post-processing
1. Initial crop (Dynamic crop)
2. Reduce background gradient and neutralize background (DynamicBackgroundExtraction)
3. Reduce background gradients more (DynamicBackgroundExtraction; normalize)
4. Neutralize background (BackgroundNeutralization
5. Set white balance (ColorCalibration; use entire image since it is all stars)

Luminance processing
6a. Extract luminance channel (ChannelExtraction)
6b. Noise reduction (ATrousWaveletTransform; mask stars)
6c. Noise reduction (MultiscaleTransform; mask stars)
6d. Nonlinear stretch (MaskedStretch)
6e. Lower black point (HistogramTransformation)
6f. Sharpen stars (Deconvolution, DynamicPSF; mask background)
6g. Compress dynamic range of stars in cluser (HDRMultiscaleTransform; protect background)

RGB processing
7a. Noise reduction (ATrousWaveletTransform; mask stars)
7b. Noise reduction ( MultiscaleTransform; mask stars)
7c. Nonlinear stretch (MaskedStretch)
7d. More noise reduction (ATrousWaveletTransform; mask stars)
7e. More noise reduction (ACDNR; mask stars)
7f. Chrominance noise reduction of large scale features (ATrousWaveletTransform; mask strong signal areas)
7g. Compress dynamic range of stars in cluser (HDRMultiscaleTransform; protect background)

LRGB processing
8. Combine sythetic luminance and RGB image (LRGBCombination)
9. Remove green noise (SCNR)
10. Sharpen (Deconvolution)
11. SCNR, remove green
12. Enrich star colors (CurvesTransformation, a and b channels of L*a*b; protect background)
13. Raise midtones (HistogramTransformation)
14. Lower black point (HistogramTransformation)
15. Brighten cluster (CurvesTransformation)
16. Sharpen (ATrousWaveletTransform)
17. Final crop to 5 x 7 aspect ratio

Comments

Author

Jason Tackett
Like

Revisions

Sky plot

Sky plot

Histogram

Globular Cluster M3, Jason Tackett

Made possible by

O'Telescope BackyardEOS