pfile
m42_ngc1977_mosaic_final
L(RGB) = 6x480s(5x240s:5x240s:5x240s)
L = 1x1bin
RGB = 2x2bin
12" R-C in rodeo, new mexico (lightbuckets.com LB-0003)
stacked with deepskystacker
initial processing with pixinsight 1.5
- normalization of ngc1977 vs. m42 data
- all subs aligned to luminance data
- rgb merge
- combined ngc1977 and m42 data with pixel math to produce a single image
- deconvolution
- histogram stretch (x10) of merged rgb data and luminance data
enfuse:
- HDR blend of all exposures generated in pixinsight
- luminance: hard mask, mean=0.54026, sigma=0.23154
- rgb: hard mask, mean=0.64026, sigma=0.23154, l-star grey projector
- had to duplicate the unstretched exposure 8 times to recover trapezium
photoshop: remove geosynchronous satellite streaks
pixinsight:
- histogram fixes and color calibration of rgb images
- histogram fixes, dark structure enhancement and atrous wavelets on luminance image
- LRGB merge
- chop composite image back into 2 separate images
- further histogram fix of ngc1977 to better match m42
hugin:
- stitch of ngc1977 and m42 images
lightroom:
- fix red/magenta saturation (pixinsight is running without color management... long story)
- crop
comments: the deconvolution is kind of bad... its heavy duty signal processing work that requires more patience than i could muster. as a result i've got some ringing and sharpening of bogus features.
it was really hard to get the two images to have the same brightness even though the exposures were the same. different nights, different amount of moon, different sky transparency all conspire to make two identically exposed images very different.
finally this is HDR so although the relative brightness between m42 and ngc1977 should be correct, the dynamic range of both have been greatly compressed. most other treatements of these objects show ngc1977 much fainter than seen here. but what is realistic when dealing with astrophotography?
m42_ngc1977_mosaic_final
L(RGB) = 6x480s(5x240s:5x240s:5x240s)
L = 1x1bin
RGB = 2x2bin
12" R-C in rodeo, new mexico (lightbuckets.com LB-0003)
stacked with deepskystacker
initial processing with pixinsight 1.5
- normalization of ngc1977 vs. m42 data
- all subs aligned to luminance data
- rgb merge
- combined ngc1977 and m42 data with pixel math to produce a single image
- deconvolution
- histogram stretch (x10) of merged rgb data and luminance data
enfuse:
- HDR blend of all exposures generated in pixinsight
- luminance: hard mask, mean=0.54026, sigma=0.23154
- rgb: hard mask, mean=0.64026, sigma=0.23154, l-star grey projector
- had to duplicate the unstretched exposure 8 times to recover trapezium
photoshop: remove geosynchronous satellite streaks
pixinsight:
- histogram fixes and color calibration of rgb images
- histogram fixes, dark structure enhancement and atrous wavelets on luminance image
- LRGB merge
- chop composite image back into 2 separate images
- further histogram fix of ngc1977 to better match m42
hugin:
- stitch of ngc1977 and m42 images
lightroom:
- fix red/magenta saturation (pixinsight is running without color management... long story)
- crop
comments: the deconvolution is kind of bad... its heavy duty signal processing work that requires more patience than i could muster. as a result i've got some ringing and sharpening of bogus features.
it was really hard to get the two images to have the same brightness even though the exposures were the same. different nights, different amount of moon, different sky transparency all conspire to make two identically exposed images very different.
finally this is HDR so although the relative brightness between m42 and ngc1977 should be correct, the dynamic range of both have been greatly compressed. most other treatements of these objects show ngc1977 much fainter than seen here. but what is realistic when dealing with astrophotography?