Rosette Nebula HOO

[JustAstrophotography]

Dual band filter on OSC

I've only recently dabbled into dual band so my knowledge is insanely limited. Thought I might share one of my latest attempts.

First image is processed normally in Startools using compose mode and dropping stacked file into red/green/blue. Under color module I chose the first preset of HOO in the list, then pushed the blue slider to (hopefully) bring out the OIII

Second image I played around with creating a starless image using heal module, I probably could have done better but for the most part I was happy with what I got.

Third image I had merged, using layer module, the starless image and a second process of the original stack where I follow the same workflow as I initially did but on the second stretch I ran it through Film Dev and only stretched it enough to faintly see the nebula but really to pull out the stars. Hopefully all that made sense

I have no idea if this is a right or wrong way of doing this but it was interesting to mess around with. I actually don't mind the little halos around stars but I know some people edit them out or at least that's what it looks like they're doing. So it was my attempt at doing that but only using Startools and not creating a starless image with Starnet ++

When I layered them, using screen filter, I felt like I lost a bit of the blue that was in the starless image.

I'm curious what other dual band filter users workflows looks like!

Details:
Redcat51
Asi 183mc pro
Asiair plus
Eq6 r pro
5 min frames/ 4 hours total
DSS
Startools

[decay]

That's a fine Rosette Nebula Seems that you captured a great bunch of Ha.

I started last summer with duoband imaging so I'm not sure if I should be the one to give comments But: I captured the same target just one or two weeks ago I dialled in a bit more OIII to bring out some more information regarding the distribution of Ha and OIII. But of course, that's just a matter of taste I guess.

I'm only using a very basic workflow with my duoband images. I order to reduce the teal star halos, I do a second run of the Color module with a star mask and saturation reduced to zero. Together with SVDecon and Shrink this often gives a decent result - at least for me But of course, I think you succeeded well reducing the teal star halos with your workflow I wonder if this complex workflow is really necessary in your case - hard to say, your dataset characteristics seem to be somehow different than mine - most probably due to fact that you do use a refractor and I a Newtonian scope.

Best regards, Dietmar.

[JustAstrophotography]

Thanks Dietmar!

It was only the third time using the duoband filter so I'm sure there's plenty more for me to learn with it. I'm going to give your workflow a spin and see how that works out for me so I do appreciate you sharing that. Plus its something I can keep in mind because I recently got a newt and I'm dying to use it.

I wonder if this workflow would be worthy of putting RGB stars into narrowband(duoband) imaging..?
I did capture 6 hours duoband and 4 hours uv/ir on California Nebula

Clear skies
Justin

[firebrand18]

Hi Justin,

Nicely done with the Rosette, looks great! Welcome to the Duo-band Bi-Color group!

It so happens I posted a similar thread last week with my Heart Nebula and have been going back-and-forth with @Mike in Rancho on processing tips; very much a matter of taste on how far you push Ha and OIII. You can see my 3 variations on the post below; first is pushed heavy on Ha, second is default Color module Ha/OIII bias, and third is H(H+O)O. Happy to offer any ideas.

viewtopic.php?f=11&t=2768

Cheers,
Nick

[Mike in Rancho]

Good Rosettes, Justin.

I think both #1 and #3 look good. Starless by itself always looks weird to me.

Nice carnival/zoo/whatever it's called there too. Hard to say if you have to go through the star heal step to get the final appearance you want. There's always fringe killer. Also, you can tame/bleach stars in Color with highlight repair as well as some settings in Shrink. And if you put iterations to zero you don't even have to literally shrink anything.

When you used compose in order to prepare for Color matrices later, did you select the bicolor from OSC option that blends the G and B and clones them identically back into the channels? I'll have to check out what that's like in 1.9 now including with our early image-type inputs, but in such bicolor cases the B and G sliders are sort of interchangeable.

[Startrek]

That’s a fine result Justin considering your limited experience with OSC Dualband processing
Some excellent detail in that Rosette
Sorry but I’m not a fan of starless images to be honest.

I don’t know if this will help or confuse you further but below is my workflow on a recent image NGC 2070 Tarantula Nebula
Just remember with processing one shoe doesn’t fit all so this example is a “guide only”
I’ve been using Startools since version 1.4 ( 5 years ago )

Workflow Startools Version 1.8 for OSC Bi colour narrowband rendition

NGC 2070 Tarantula Nebula using 6” f6 GSO Newtonian Reflector ZWO2600MC camera cooled to -10C , Gain 100 , offset 50
Antlia 2” ALPT Dualband OSC filter
140 x 120sec dithered subs , 40 x Flats and 60 x Bias Calibration
Stacked in DSS ( Multiple Nights using reference frame)

Open Startools

Loading Data Set
Select “Compose” at top
Select Luminance Color L + Synthetic L from R (2 x G)B, R (GB) (GB) (Bi Color from OSC / DSLR )
Load data set into Red/Sii
Load data set into Green /Ha
Load data set into Blue / Oiii
Press Keep
Then select Linear

Notes: Use Before and After buttons top right to review changes and results when adjustments are made in a module or mask
Use Zoom in and Zoom out button , top right to inspect changes and results when adjustments are made in a module or mask

Start Processing

Select “AutoDev module” to inspect image for visual anomalies and other problems
Press Keep

Select “Bin module”then choose between 60% to 70% using slider at bottom depending on your image scale and seeing conditions. Startools Binning is scalable and achieves best results with oversampled data, otherwise Stars become “blocky” when under sampled.
If unsure don’t Bin your data and see how your Stars looks after a global stretch. Binning provides an improvement in SNR at the expense of some loss of resolution which may not be seen or apparent in any case.

Select “Crop module”
Draw a region around image to Crop edges for stacking artefacts or Crop out unwanted part of image ( eg waisted interstellar background )
Then press Keep

Select “Wipe module”
Use default Basic Wipe
Increase Dark Anomaly Filter to 3 or 4 pixels
Select Keep , then Ok , then Keep again

Select “AutoDev module” again
Press Redo Global stretch
Increase Ignore Fine Detail to 3.5 to 4.5 pixels to reduce unwanted noise or fine grain
Draw a Region of Interest ROI around main structure or object to bring out detail in this area whilst keeping a lid on brightness and background noise. Adjust size and shape of ROI until happy with final Global Stretch of image
You can also cancel ROI if image has a satisfactory level of contrast , detail , brightness and noise floor
Detector Gamma can be increased from default 1.00 to 1.05 or higher to increase brightness
Shadow linearity can be increased from default 50% to 60% or higher to supplement the increased Gamma
Once happy with final global stretch press Keep

Select “Contrast Module”
Adjust Locality from default 50% to 40% or 30% to locate contrast in the main body of the object
Adjust Shadow Dynamic Range from default 50% to 40% or 30% to reduce strength of contrast amount
Once happy with Contrast press Keep

Select “HDR module”
Use Tame preset at top to reduce brightness in core of Nebula
Adjust Gamma Highlight from default 3.00 down to 1.10 or up to 1.20 as required
Once happy press Keep

Select HDR module” again
Use either Reveal or Optimise to enhance detail in main structures
Adjust both Highlights and Shadows Detail Boost as required (Usually 20% to 30% is sufficient )
Once happy with final detail press Keep

Select “Sharp module”
Press Auto generate mask
Default settings work pretty good
Press Keep

Select Mask , Clear then Keep

How to Tame the bright core area of the Tarantula Nebula -
Press Mask at top again
Make sure “Brush mode”at bottom is set to Flood fill Lighter pixels
Now click on bright core centre and adjust mask area using shrink and grow buttons. If too aggressive clear Mask and try again
Alternatively you can use “Brush mode” Lasso and draw your Mask manually
Once Mask has covered super bright nebulosity areas press Keep

Select “Shrink module” to tame the super bright core area
Press button Keep Mask as is
Increase iterations to 25
Choose Dim mode
Increase Un glow strength to 30%
This should dim the super bright core. Adjust above as required to achieve required result
Once happy press Keep

Select Mask again , press Clear the Keep

Select “SV Decon module” Spatially Variant Deconvolution
Press Auto generate extra sensitive mask
Once Mask is set press Sampling button at top ( image will turn Red ready for selecting Star samples )
Zoom in to 300% and starting from far left or far right hand side of image select Star samples that have either a tiny yellow or green core. Samples should be selected across the whole image ( minimum of 10 samples , maximum of 20 samples) Avoid object cores and heavy nebulosity areas as samples
Also select only single stars not stars than a join other stars
Once samples are all selected and image has timed out ready , then press Result button a top
To increase level of Deconvolution , increase iterations from default 10x to say 12x or 14x however this may increase ringing artefacts around stars. To minimise ringing artefacts around stars , increase De ringing amount to 0.90 and De ringing Detect to 80% , this should help reduce de ringing
Once happy with results, press Keep at top

Select Mask again and press Clear then Keep

Select “Super Structure module”
Set Full Mask
Press “Isolate” preset and only adjust Strength to 20% to 25%
This tones down the brightness in the main structures, background and Star field to expose more fine detail
Press Keep when satisfied with result

Select “Colour Module”
Press Fill Mask
Go to Matrix at bottom and select your SHO colour blend mapping ( 4th one from the left , SHO 60Sii + 40Ha, 30Sii + 30Ha + 40Oiii, 100Oiii )
Style is Scientific ( Colour Constancy )
Saturation Amount 230% to 250%
Bright Saturation Full
Dark Saturation 4.0 to 5.0
Cap Green 0%
Highlight Repair 4 to 6 pixels
Red Bias Reduce 1.00
Green Bias Reduce 1.50
Blue Bias Reduce 1.05

Once happy with Colour of image press Keep

Select Superstructure module
Leave as default Dim Small preset
Adjust Strength to between 50% to 100% as required.
Brightness Retention select Off.
This should reduce noise and unwanted brightness in main structures including background and Star field
Once happy with result press Keep

Select Mask again , press Clear then Keep

Finally select Track/ Noise module
Press Apply noise reduction
Use default ( usually 4.0 to 5.0 pixels ) or increase Grain Size to smooth out noise grain
Walking noise can be left to default 1.0 pixels unless you have noticeable walking noise in your image , then increase as required
Press Next
Once your happy with final noise reduction press Keep

Select Save at top to save your completed image to nominated folder and file

Processed image completed

Attached final image ( original image size was 1.3 MB ) downsized to 500KB in Startools forum

Clear Skies
Martin

[JustAstrophotography]

Hi Justin,

Nicely done with the Rosette, looks great! Welcome to the Duo-band Bi-Color group!

It so happens I posted a similar thread last week with my Heart Nebula and have been going back-and-forth with @Mike in Rancho on processing tips; very much a matter of taste on how far you push Ha and OIII. You can see my 3 variations on the post below; first is pushed heavy on Ha, second is default Color module Ha/OIII bias, and third is H(H+O)O. Happy to offer any ideas.

viewtopic.php?f=11&t=2768

Cheers,
Nick

Thanks Nick!
Yes, the Dou-band certainly opens up some doors for me with capturing DSO. I did read through your thread and I got some ideas that I want to try out. Your Heart Nebula looks amazing too!

Good Rosettes, Justin.

I think both #1 and #3 look good. Starless by itself always looks weird to me.

Nice carnival/zoo/whatever it's called there too. Hard to say if you have to go through the star heal step to get the final appearance you want. There's always fringe killer. Also, you can tame/bleach stars in Color with highlight repair as well as some settings in Shrink. And if you put iterations to zero you don't even have to literally shrink anything.

When you used compose in order to prepare for Color matrices later, did you select the bicolor from OSC option that blends the G and B and clones them identically back into the channels? I'll have to check out what that's like in 1.9 now including with our early image-type inputs, but in such bicolor cases the B and G sliders are sort of interchangeable.

I agree with the starless images. Its never my go to for display, in this case it was only my stepping stone to get to a final image, which maybe I don't have to do based on other workflows for reducing halos. I'm going to try the second shrink with iterations set to zero to repair stars!
In compose I do select the Bi-color " L+Synthetic L from R(2xG)B, R(GB)(GB), and I set green channel to the twice exposure to red and blue (is this correct, I feel like I read that somewhere). I have not used 1.9 yet but maybe I should mess around with it.


@Startrek
Thanks for the detailed workflow!
It looks like our workflow is pretty close with a few miner differences. And your Tarantula Nebula looks absolutely awesome! On shrink module I've always stuck with default on iterations but it looks like you're all the way up to 25. I'll give that a spin too.

[Startrek]

Justin,
Please be aware that using “Shrink” in this unorthodox way is only for Nebula with super bright cores like M42 and this Tarantula nebula only !!!!
Just to point out that I used “Shrink” with a specific localised Mask only to Tame the central core region of this Nebula as it’s super bright ( Not for general Star field size reduction which it is normally used for ) I did note the reason in my Workflow. This is an unorthodox way of using Shrink but it worked well to dim down just the core region using such a high iteration setting of 25.

For general Star field, Star reduction I used a Star Mask with iterations set at default or 10 and the Tighten preset

Clear Skies
Martin

[JustAstrophotography]

@Startrek
Ah, ok!
I think I either missed that entire section explaining it or I just wasn't paying attention.
Just curious, does tame in HDR just not do what you want it to/ or not enough of what you're looking for?

[Startrek]

Justin,
Again read my workflow regarding HDR , yes. I did use it and it adjusted the brightness of the main structure or body of this Nebula perfectly, but it cannot dim super mega bright cores in this Nebula and M42 for example

Clear Skies
Martin