Many thanks to Leandro for capturing the data and producing the Ha and OIII stacks.
@Leandro from Argentina captured the data during April/May 2022, extracted the Ha & OIII, performed calibration and integration to produce an Ha and OIII stack. I (@turfpit from the UK) blended the narrowband stacks using PixInsight.
Ha_OIII__final_scaled.jpg (823.69 KiB) Viewed 1092 times
Equipment & Capture (Argentina)
Montura NQH 3 Orion GoTo System
Telescope Sky Watcher 200 x 1000 Reflector Newton
Filter Optolong L-eNhnace 1,25"
Camera Altair Astro 183C Pro
Guiding SkyWatcher Evo 50 x 240 mm, camera Altair 130 M, PHD2
Register the OIII stack to the Ha stack using StarAlignment
DynamicCrop to ensure 2 identical size images using New Instance
DynamicBackgroundExtraction on AutoStretched (boosted) images using manually placed points in regions with no nebulosity/stars. About 10 points placed. Same settings applied to both images via New Instance
PixelMath to blend the 2 resultant images
Apply gentle CurvesTransformation to result
SCNR to remove slight green caste
PixelMath.JPG (26.43 KiB) Viewed 1092 times
These are the stacks before blending:
Ha_OIII_images_before_blending.JPG (119.59 KiB) Viewed 1092 times
I am working through PixInsight and am nearly at the end of the 6 week trial license. I expect there will be plenty of improvement to be had with experience - any suggestions welcome here.
What a good idea and a very nice collaborative image. You have got a long way with PI processing quickly... it is fun to play with the Pixmath recipe in NB OOH images and Pixmath is a great tool imo.
I found - and doubtless you are noting the same - that with most type II HA/ OIII objects the HA dominates and if you stretch both the OIII and HA linear images equally you do end up with very red images. ...as per your equal stretch HOO image which looks pretty good (and honest). Most astrophotographers don't seem to do that though and, at the cost of more noise, go for mixing HA and OIII images that are stretched up so as to be about equal to eachother - and I guess that in your final image that is where the mixed OIII/ HA regions have come up white?
The only things that I can think of to try that might be different depending on taste ? --
1) Try using LRGB combination (just L and chrominance noise reduction clicked) to swap the HA luminance into the final image - then curves to readjust luminance and saturation. It should sharpen up the clarity of the nebula outline and nicely reduce the stars. it seems a reasonable thing to do since nearly all of the signal is HA -nowhere is OIII dominant - and doing this helps suppress the blurring effect of the OIII image noise while still preserving the OIII effect on balancing chrominance.
2) OR different approach only requiring a little more work? First stretch the OIII and HA images separately -- using histogram transformation. Then make each starless using the starnet tool . Then combine the two images and apply curves / noise reduction -MLT say - to the starless image ...then add stars back in again. With this approach (if you have the data) you can put RGB natural colour stars back into the picture. But it would need some (e.g) OSC RGB data to do that.
PixInsight has been a hard six weeks. Once I got to grips with how the software interface works, progress was fairly rapid. The 'New Instance' saved to the desktop makes life easier when working with multiple images such as in LRGB and NB processing. I used some M31 data most of the time. heavy handed use of automation with DynamicBackgroundExtraction left me with 'black holes' surrounding the bright objects. Once I used hand placed reference points in DBE on an Autostretched (boosted) image the problem went away. The biggest eye opener has been masks/inverted masks - the ability to apply noise reduction to the background only and apply sharpening to the object only certainly helps with post-processing. I have been able to produce finished images using only the PixInsight software.
Thanks for the info in 1) and 2) I am experimenting with these at the moment. I agree with you that application of OIII can have a detrimental effect on results for this object.
Personally, I like the monochrome images, particularly the Ha.
The PI trial license is due to expire in day so there is a decision to be made.
This is the LRGB (OSC) image from the Altair Astro 183CPro camera (240x30s @ gain=400). Retaining detail at the bright centre was the hardest part of processing:
LRGB-scaled.jpg (869.32 KiB) Viewed 967 times
The colours seem reasonably balanced (without having to torture any pixels):
LRGB-histogram.JPG (64.24 KiB) Viewed 967 times
Thanks again to Leandro for capturing the data and carrying out the calibration/integration of the OSC data and the extraction/calibration/integration of the Ha & OIII data.
For me the "red version" is the best one. But that is personal too I guess, for me there is more detail to be seen in the red contrasts.
Question: do you use software like Starnet or Starxterminator to isolate the stars so that those are not stretched? It gives you way more "control" over processing the nebula without blowing out the stars.
I like the 'red' one as well - a simple blending of Ha (R) and OIII (G & B).
Starnet is one of the processes available in PixInsight (as well as being standalone software). I have just spent 6 weeks working with a PI trial license. There are a lot of processes (and documentation) to get to grips with. I can see how 'starless' would help processing big galaxies or nebulae where there are bright stars present.
Dave
PixInsight 'All Processes':
PixInsight_Processes.JPG (126.32 KiB) Viewed 952 times