Lots of horrible Elephant’s trunk data collected this last summer while fixing the F4 Newtonian
1) 9h of HA data -- poor collimation due to mirror shift (too loose) leading to elongated stars (r ~ 0.83) but otherwise sharp/ good SNR
2) 1h of RGB data -- nasty ~ trefoil star shapes due to mirror clamps being too tight
But also some good data
3) 6h of O3 data -- nice almost round stars (r ~ 0.94)
Somewhat surprisingly - it was possible to recover a good RGB OOH image from this this dog’s dinner of data. Encouragement perhaps not to simply discard flawed data ? - because realistically - it is usually flawed in just one aspect or another that good data can sometimes fix.
In outline and in case useful - the strategy that seemed to work to produce a good RGB stars/ OOH image was the following ...
1) HA data. Use deconvolution to slightly sharpen up the HA image (the r= 0.83 stars were poor but at least consistently so and thus it was possible to derive a representative PSF) and afterwards remove the stars (starExterminator 2.03).
2) RGB data. Extract the badly-shaped but nevertheless faithfully coloured RGB star set (StarExtractor) and then fix the star shape by transferring over the luminance from the star set extracted from the good O3 data. So then had a set of RGB coloured and good shape stars.
3) Combine the starless O3 and HA data into an NB image - apply HA luminance (since it dominates and is more detailed) - stretch (exponential and local histogram equalisation) and then add RGB stars.
Images below show the final image, the distorted star shapes of the component images and the subtle but still useful effect of applying deconvolution to the HA image. Lastly an image of the same object compiled a year ago in 2021 compared to this year's effort because I was seeking assurance that I am making progress
The full resolution image as well as capture details are linked here.
https://www.astrobin.com/kqyufn/0/
SC capture using the VX12 F4 12 inch Newt was as usual , at 0.81 arcsec/ pixel and under Bortle 6 skies over a number of nights
Tim
Rescuing the Elephant's trunk from a dog's dinner
Forum rules
Please upload large images to photo sharing sites (flickr, etc) rather than trying to upload them as forum attachments.
Please share the equipment used and if possible camera settings to help others.
Please upload large images to photo sharing sites (flickr, etc) rather than trying to upload them as forum attachments.
Please share the equipment used and if possible camera settings to help others.
Rescuing the Elephant's trunk from a dog's dinner
- Attachments
-
- IC1396improve.JPG (76.85 KiB) Viewed 693 times
-
- IC1396_starshapes.JPG (90.69 KiB) Viewed 693 times
-
- IC1396_dECON.JPG (61.35 KiB) Viewed 693 times
-
- IC1396_etrunk+NBimage_StarXless_LRGB_starlessHAlum_EXP_LHE1.2_curves_MLT_PIXM blend with RGB stars with O3 luminance_affinity16bit_PX_BIG_mini.jpg (964.04 KiB) Viewed 693 times
- admin
- Site Admin
- Posts: 13350
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: Rescuing the Elephant's trunk from a dog's dinner
Hi Tim,
once again great images and a nice write up of how you achieved them. I rarely get around to processing images that I capture, let alone work on them to this level of detail. Once the image is off the camera and onto disk then my job is done
cheers,
Robin
once again great images and a nice write up of how you achieved them. I rarely get around to processing images that I capture, let alone work on them to this level of detail. Once the image is off the camera and onto disk then my job is done
cheers,
Robin
Re: Rescuing the Elephant's trunk from a dog's dinner
Tim
An interesting post - a dose of realism instead of the illusion that everything always works. This should encourage others that a failure is part of the learning process - it is how we really progress. The deconvolution before and after comparison demonstrates the value of that particular process (once it has been understood how to use the PSF).
From now on I will always think of this as the Dog's Dinner Nebula.
Dave
An interesting post - a dose of realism instead of the illusion that everything always works. This should encourage others that a failure is part of the learning process - it is how we really progress. The deconvolution before and after comparison demonstrates the value of that particular process (once it has been understood how to use the PSF).
From now on I will always think of this as the Dog's Dinner Nebula.
Dave
Re: Rescuing the Elephant's trunk from a dog's dinner
Dave, Tim,
a reminder that we only really learn from our mistakes, not when things go as expected!
Great explanations, upvotes to both
Regards,
Carl
a reminder that we only really learn from our mistakes, not when things go as expected!
Great explanations, upvotes to both
Regards,
Carl
Re: Rescuing the Elephant's trunk from a dog's dinner
Robin, Dave, Carl - thanks indeed for the kind comments.
Particularly at small image scales and with large reflectors, wind, clouds, dewing up, rats eating cables etc etc it does seem a bit self-defeating to set too high a standard for what is acceptable during data acquisition. Certainly good optimisation of the optics is critical for fine planetary and lunar work but maybe for deep sky nebulae etc the detail is usually on a scale that is more forgiving and, in practice, the final image quality not so dependent. At least that is what I think I am learning at the moment ... but maybe a different view 6 months down the road
Processing tools are becoming so good now - plus one's understanding of them grows - that there also seems to be a lot to gain from just reworking old data (cf Menno's Heron galaxy too).
Tim
Particularly at small image scales and with large reflectors, wind, clouds, dewing up, rats eating cables etc etc it does seem a bit self-defeating to set too high a standard for what is acceptable during data acquisition. Certainly good optimisation of the optics is critical for fine planetary and lunar work but maybe for deep sky nebulae etc the detail is usually on a scale that is more forgiving and, in practice, the final image quality not so dependent. At least that is what I think I am learning at the moment ... but maybe a different view 6 months down the road
Processing tools are becoming so good now - plus one's understanding of them grows - that there also seems to be a lot to gain from just reworking old data (cf Menno's Heron galaxy too).
Tim