You need to log in to create posts and topics.

M81 Wide Field, Rokinon 135 f2

M81/82 and a bunch of other stuff with IFN galore.  This was taken with a selection of data from recent Rokinon 135 f2 shots from December, January, and also test data from spring 2016, since it was similar framing.

2018-2019 data:

Ha: 29@300s

R: 32@300s

G: 40@300s

B: 29@300s

L: 37@300s

Spring 2016:

R: 28@300s

G: 28@300s

B: 29@300s

L: 88@300s

I was a little hesistant to see what improvements would be coming from stacking data from 2016 to recent, so I stacked files twice and registered/cropped the separate stacks to the same region.  I then used SubframeSelector to measure SNR from each comparative stack.  Then I weighted both stacks with their respective SNRs and averaged both together with weighting using PixelMath and ensured the final SNR improved.  The additional luminance helped the most...really brought up the IFN.

Uploaded files:
  • M81_LHaRGB_v1.jpg

Wow, IFN was brought out very nicely plus very lovely galaxies. However, my processing based on 2018 data didn't show the expected IFN. Is this mainly due to less of 70@300 luminance, or it's due to bad processing? The image is posted at

I checked the data, and noticed that after ABE and/or DBE, the background revealed a prominent shape and I don't see a way to pull up the sharply edged IFN bends. It doesn't seem to be the additional L data.

There was more than doubling in my measurements after using the test data BUT I had stacked the 2018 and 2016 data separately (using the same close-to-best FHWM registration file.)  I just recently joined so I'm not sure what the conditions were from the 2016 data, but the vignetting looked far less than the 2018 data (stopped down to f2.8?), so while I used the master flats, darks, and bias on the 2018 data I DIDN'T use the flat on the 2016 data.  Then, stacking them seperately I cropped to the same overlapping area, performed DBE on the files separately (since vignetting, sky noise were different between the 2), then measured SNR.  For an experiment now, I just did the overlap area crop on the files then performed ABE with default settings (using the divider operation for better vignetting association.)  I then did SubframeSelector on both, and found the SNR.  Using PixelMath I made the following weighted average with:


Saving the file, I now show the SubframeSelector result with 1 being the combo file, 2 the 2018 data, and 3 the 2016 data.  Note the basic doubling of SNR from either.

The next picture shows the 2018 data file on the left, combo file on the right with just a MaskedStretch at 0.28.  Observable higher contrast in the combo file.  So, it's definitely worth it.  Though SNR from individual frames are about half the SNR from 2018 data, again indicating a stopped down nature.  Color files thus did not have as much improvement as the luminance but it was still there.

In my processing I shrank stars and highly increased contrast cause that's how I roll, but even now I'm reconsidering my processing.  I'm never happy with my end results.

Edit:  Holy cow you got a TON more resolution on M81 and M82 than I did!  Did you combine with some longer focal length data on those?  It looks nice.

Uploaded files:
  • SubframeSelectorM81.JPG
  • M81Compare.JPG

Phew, liking THIS version a lot better.  Less dramatic on the IFN, more color, etc.  Most of the image is LRGB, no Ha inclusion at all.  Then I developed with the Ha, only selected the M81/82 area to paste the red colors for the emphasis.  Allowed me to keep color throughout the image for the other areas without it turning funky.

Uploaded files:
  • M81_LHaRGB_v2.jpg

Is 2016 data in the test data? I didn't see it. Was it in the 2016 Q3/Q4 production data?

Is 2016 data in the test data? I didn't see it. Was it in the 2016 Q3/Q4 production data?

Okay, finally, I managed to bring IFN up: by LHE.