Cookie consent

AstroBin saves small pieces of text information (cookies) on your device in order to deliver better content and for statistical purposes. You can disable the usage of cookies by changing the settings of your browser. By browsing AstroBin without changing the browser settings, you grant us permission to store that information on your device.

I agree
Contains:  M 92, NGC 6341
Getting plate-solving status, please wait...
M92, Globular Cluster in Hercules, 


            Jason Tackett
M92, Globular Cluster in Hercules

M92, Globular Cluster in Hercules

Technical card

Resolution: 3505x2503

Dates:May 22, 2015May 23, 2015

40x120" ISO400
37x240" ISO400

Integration: 3.8 hours

Darks: ~20

Flats: ~30

Bias: ~50

Avg. Moon age: 4.89 days

Avg. Moon phase: 24.83% job: 665689

RA center: 259.292 degrees

DEC center: 43.137 degrees

Orientation: 78.733 degrees

Field radius: 0.556 degrees

Locations: Home base, Yorktown, VA, United States


This image of M92 is an HDR composite of two sets of subframes at 120s and 240s. In order to reduce the background noise as much as possible, both sets of calibrated subframes were averaged together. Next, I used HDRComposition with a binarizing threshold of 0.01 to replace the brightest stars and the stars in the core of the globular cluster with those from the 120s set.

Revision B notes:
The main difference with Revision B is that I discovered that I have been mishandling my flat frames during the calibration process which skewed my colors to the blue. My flat box has a red hue and I never realized that the blue channel was gravely underexposed in my flat frames. As a result, my light frames ended up with a substantial blue shift. In the case of this image, the blue was so severe that no manner of manipulation could bring out the yellow stars in the cluster. This was probably the reason why my M33 image looked monochrome blue as well. To save my master flat image, I balanced the color channels by using SplitCFA to extract the CFA components and then scale the green and blue channels to the level of the red channel using PixelMath (expression: $T*(median(CFA0)/median($T), applied to CFA{1-3} ) I then recombined the CFA components using MergeCFA and calibrated as usual. The difference is substantial! There wasn’t a color shift in my calibrated light frame and highlight colors were preserved. The downside is that I amplified the noise in the underexposed channels which showed up as a noisy red cast to the background on the left and right where my vignetting is worst. DynamicBackgroundExtraction and cropping helped with mend this issue. In the future (until I get a new flat box), I plan to take at least two sets of flat frames per target - one ensuring R is properly exposed and one ensuring G and B are properly exposed.

Processing Workflow (PixInsight)
Revision B changes are in italic. Removed/altered processes from the original image are struck out.

1. Calibrate light frames using darks, bias, flats
2. Create two sets of master light frames: one for the 120s set and a second for the 120s & 240s set (ImageIntegration)

For the two master light frames;
1. Initial crop (Dynamic crop)
2. Reduce background gradient and neutralize background (DynamicBackgroundExtraction, division subtraction)
4. Neutralize background (BackgroundNeutralization)
5. Set white balance (ColorCalibration; use entire image since it is all stars)

Single HDR image processing:
6. Combine two master light frames into a single 64 bit light frame, so that that stars from the 120s master light are used in the core of the cluster (HDRComposition; binarizing threshold 0.01)
7. Set luminance coefficients to 0.33333 (RGBWorkingSpace)
8. Non-linear stretch (MaskedStretch script; target median background 0.15, clipping fraction 0.01)
9. Reduce background brightness; increase contrast (CurvesTransformation to RGB/K)
10. Reduce background luminance noise (ACDNR to luminance with luminance mask)
11. Reduce background chrominance noise (ACDNR to chrominance with luminance mask)
12. Increase star color saturation (CurvesTransformation to Saturation with range mask)
13. Increase star color saturation some more (CurvesTransformation to Saturation with range mask)
14. Reduce background brightness slightly (CurvesTransformation to RGB/K)
15. Reduce green (SCNR, amount 0.6 1.0)
16.Sharpen stars slightly (MultiScaleMedianTransform, bias +0.01 to layer 3 of 5, target Lightness Unsharp Mask, StdDev 0.8, Amount 0.6)
17. Remove dark pixels (PixelMath [Schwarz script]; protect stars)
18. Cosmetic correction to reduce purple cores in a few over-exposed stars. Using a smooth range mask selecting just the offending star cores, first brighten cores close to white (InterchannelCurves; Target CIE L, Reference CIE L; increase saturation) then reduce purple (InterchannelCurves, Target CIE c, Reference CIE L, reduce saturation).
18. Lower black point for each RGB channel separately until median background value is around 0.1 (HistogramTransformation BackgroundNeutralization, Target Background 0.101)
19. Shift greens down slightly (HistogramTransformation, lower blackpoint of green channel)
20. Final Crop to 5 x 7 aspect ratio, rotating so line of bright stars form an arc from the upper left corner to below M92 to add weight to the composition (DynamicCrop)
21. Set ICC profile to sRGB for web publishing (ICCProfileTransformation).

Manfred Schwarz’ galaxy tutorial:



Jason Tackett
License: Attribution Creative Commons


  • M92, Globular Cluster in Hercules, 


            Jason Tackett
  • Final
    M92, Globular Cluster in Hercules, 


            Jason Tackett

Sky plot

Sky plot


M92, Globular Cluster in Hercules, 


            Jason Tackett

Made possible by

O'Telescope BackyardEOS