First nebula I've tried since returning to the hobby, I'm not happy with the ringing on the larger stars and I'm underselling the data I have. Having not used Pixinsight for a while I've watched a few walkthroughs and done some research however I'm sure there are better processes and scripts I can use to bring out the details I'd love hear any to investigate. Previous I used to be too agressive with my processing so I'm trying to rain it in, but I think I've lent too far the other way now.
Dates: 06 - 08, 13-14 Feb
Acquisition: 106 x 300s = 8.8 hours Gain 0 Offset 50 Temp -10C
Equipment: WO FLT110 FF/Reducer F6.3, ASI2600MC Pro
Pixinsight:
Pixinsight processing steps:- Image Calibration
- Cosmetic Correction
- Debayer
- Subframe selector
- Star Alignment all days together
- Local Normalization
- Image Integration
- DBE (Tol 1.6, shad 4, size 150) - subtraction , Division
- photocolourcorrection
- RGBWorkingSpace (set values to 1)
Create luminance layer- STF Standard
- StarNet2 (256, Create starmask, 2x upsample, linear data)
- Clean starless image with clonestamp
Starless image- Stretch until bright parts are visible, then final stretch moving black point
- rename as mask and invert
- MSLT (4 layers 3,0.5,3/2,0.5,2/1,0.5,2/0.5,0.5,1) PSFImage (With STF) Sens -0.7
- Clone
- 4 further slight streches to the starless image
Star image- Deconvulution - External PSFImage
- On the stars only luminance image, I did four slight streches with HistogramTransformation, with just the Midtone stretched to 25%
Combining Lum- Used PixelMath to recombine the Luminance stars / starless images using: ~((~starless)*(~stars)) and rename "Lum"
- Create a clone of the resulting Luminance image, then apply a mask around bright cores and apply HDRMultiscaleTransform and then LocalHistogramTransformation.
- Use PixelMath to combine the "Lum_hdrmt" and "Lum" images - in this case with 0.5*Lum + 0.5*Lum_hdrmt and rename "L50"
RGB- Clone
- Stretch until central part is visible, then final stretch moving black point
- rename as mask and invert
- MSLT (4 layers 3,0.5,3/2,0.5,2/1,0.5,2/0.5,0.5,1)
- Repaired HSV Separation (script)
- Clone "RGB_Repair" and apply MaskedStretch & rename "RGB_MS"
LRGB- On RGB_MS i then ran ChannelExtraction to create HSI components. Delete the "I" image
- use ChannelCombination with HSI selected, using the H & S images created from the RGB_MS file and the "L50" as the "I" component. Rename the result as "HSI"
- Extract the luminance from the HSI file and using ScreenTransferFunction / HistogramTransformation create a high contrast image to use as a mask to adjust the colour saturation (rename: Sat_mask)
- Apply "SatMask" to the HSI image & use Curvestransformation to increase saturation. (Two adjustments were made)
- Apply SCNR with a small reduction to the Green channel (reduced to 0.9 x original)
(http://gallery.orpington-astronomy.org.uk/albums/userpics/10056/normal_M78_v2_1_small.jpg)
Ivor,
I will be interested to see the expert comments on this. That is an amazing amount of processing - way more than I have ever applied.
FWIW I have never been very happy with using Starnet - for me I find it gives a muddy, fuzzy blob. Obviously, it works much better for you. I always use it on a stretched image which may be my major mistake.
I am also lazy and I have completely turned to the WBPP script - I find it does better than when I try to use the individual stages.
Star fringing issues aside it's a nice image and the dust clouds are very clear.
Regards,
John
A nice result Ivor. An amazing amount of processing. I don't use Pixinsight so cannot comment at all.
I personally would have increased the colour saturation as it does look rather anaemic at the moment.
You say this is a work in progress. What is your total imaging time so far?
Also to say that this target benefits from the addition if Ha, especially with Barnard's loop so close, you may have already planned to do this.
Carole
Quote from: JohnH on Feb 22, 2023, 21:22:42FWIW I have never been very happy with using Starnet - for me I find it gives a muddy, fuzzy blob. Obviously, it works much better for you. I always use it on a stretched image which may be my major mistake.
It is the first time I've used starnet2 and the results were pretty good with the settings above, however my cleaning up with clone stamp wasn't as effective as I'd like so I'm researching alternative approaches.
Quote from: Carole on Feb 22, 2023, 23:22:45I personally would have increased the colour saturation as it does look rather anaemic at the moment.
You say this is a work in progress. What is your total imaging time so far?
Also to say that this target benefits from the addition if Ha, especially with Barnard's loop so close, you may have already planned to do this.
I tried a few different saturation levels using Curvetransformation and colorsaturation however none gave it the punch I'd hope, which makes me think my processing approach is wrong.
My original mono camera was a SBig ST8300 which is now obsolete so I bought the ASI2600MC so I don't have a way to collect Ha right now. I'm considering getting one of the extreme filters to resolve this however I'm also considering replacing my FLT110 with a faster scope which will impact the choice of filter.
Total RGB data is 8.5hrs so I don't need anymore.
You both commented on the amount of processing which makes me wonder how much time do people spend on processing?
I estimate the above took me 4-5 hours, however once I'm familiar with Pixinsight again I'll be down to 3 hours. How long do others spend on an image?
Regarding it being a work in progress, I know my data is good however my lack of skill in the processing is letting me down. I always look at the images from the likes of Mark and remind myself my lack of finesse with the processing tools is letting me down so I keep wanting to revisit the end result.
I do spend several hours processing but some if this is combining several filters and data from a dual rig.
But if it gets the results you want, who cares how long you spend on it.
Carole
My Cone, Xmas, Foxfur pic took about 1 hour to stack with drizzle in WBPP. I wouldn't bother doing all the manual steps now they've been automated so well. Then maybe 2 hours for SCC, ABE, (quick), EZdecon, EZdenoise. I keep the starless version and masks generated for later use. Then I try different stretch methods to see what looks best before EZStarReduction, curves and saturation tweaking. Then into Affinity as a tif for selective tweaks, merging starless data if appropriate, masked Clarity sharpening etc.
About 4 hours for that image all told.
Very good Ivor! Thank you for documenting your processing; that's a lot of work in itself. M78 is not an easy object and you have a very good image there.
Roberto
Just had a look at the Gallery image, very nice subtle processing showing the different stuff. Lovely Ivor.
Thanks for all the feedback, still a lot to improve but thats part of the challenge. Before I look into WBPP process I want to understand the underlying processes and how they can be fine tuned. The subselection functionality is new for me and I'm getting familiar with the weighting as a number of decent subs would have got dropped if I'd gone for the default settings.
I have Affinity photo as well I'll be interested in playing with the astro functionalty in there as well once I get a chance.
I hadn;t realise it was on the gallery, much easier to see. There is some very nice fine detail in this image, so you are going in the right direction. Pity about no Ha filter though.
Carole
I've invested in Adam Block's tutorials, I'm very impressed with what is provided. Still got lot's to learn but I think this is an improvement on my first attempt.
(https://cdn.astrobin.com/thumbs/3Z-FghjhItTX_16536x0_ieTZ0INm.jpg)
That's looking really nice now Ivor.
Carole
Hi Ivor, what's your Bortle scale (for members images report).
Carole
It is supposed to be 4 but I think it is probably closer to 5 thanks to the led lights in Tunbridge Wells industrial estate
That's a great result Ivor!
Nice image Ivor.
Clear skies
GW