Jump to content

Stargazers Lounge Uses Cookies

Like most websites, SGL uses cookies in order to deliver a secure, personalised service, to provide social media functions and to analyse our traffic. Continued use of SGL indicates your acceptance of our cookie policy.

stargazine_ep33_banner.thumb.jpg.75d09b4b1b4e5bdb1393e57ce45e6a32.jpg

Grant

IKO - M17 - Processing Competition

Recommended Posts

This one is going to be a very tough one to judge!

I'm loving the different approaches to processing this, I think it's the most varied yet! :)

 

Share this post


Link to post
Share on other sites

Oh awesome totally gonna get in on this I think ūüôā will download once I'm home¬†

Share this post


Link to post
Share on other sites

Here'my attempt. Processed mostly in PixInsight as SHO, with Ha as Luminance, then colour balance tweaked in Photoshop.

Bill

M17.jpg

  • Like 2

Share this post


Link to post
Share on other sites

Here are a couple of versions, top one with Ha-red, S2-green and O3-blue. I also used a superluminance (all three channels added with Pixelmath) and used HDR to bring out the structure in the brighter sections of the nebula.

2nd version is Hubble palate which accentuates the nebulosity in different places. Also uses the superluminance.

I need more data to process and then I'll get better. I've not imaged since mid September! Still need to finish my North America Nebula.

 

M17HaS2O3_final.png

M17S2HaO3_final.png

  • Like 6

Share this post


Link to post
Share on other sites

individual channels selectively stretched and then RGB combined in PSP (HSO = RGB).  Image cropped and rotated to match actual sky orientation.  Curves adjusted to bring out the red in the nebulosity because that's my own preference: predominantly green or blue emission regions just don't seem "right" to me, but then I am an old fossil... 

1493903374_M17-2jpg.thumb.jpg.8865adfa0878d5dfba016ae015f3d98c.jpg

 

I tried to keep the processing simple and was keen to try and preserve the "swan" at the heart of the main nebula. There is a wealth of data in that central region but I resisted the temptation to over-sharpen the nebulosity as I wanted to retain the softness from visual observations I have made of this summer beauty. 

The data was so good it didn't really need much else doing to it. Perfectly round, tight stars across the field, so any form of star reduction was simply not required. Almost zero noise or gradients, so very little noise reduction needed (a very mild edge-preserving smooth used at the end).  

Think I might sell all my gear and just play with IKI data instead!

Edited by Hallingskies
40% Ha luminance layer added to bring out more detail
  • Like 1

Share this post


Link to post
Share on other sites

Amazing datasheet! so many details... thanks again for let us play with it! 

not sure if I do justice to it, but here it goes my first attempt ! :D

Processed in StarTools 1.7 430, and Lightroom

 

03112020-M17-SGL.jpg

  • Like 1

Share this post


Link to post
Share on other sites

A slightly softer rendition; HaSHO but processed via FITS with Liberator and then the RGB combined made starless with starnet and slightly stretched. Selective colour, then Ha as Luminance. A bit of contrast &  sharpening layers plus some vibrance & saturation for good measure.

omega_s 22.png

  • Like 1

Share this post


Link to post
Share on other sites

Here is another one this time using a combination of mainly PixInsight and finished with Photoshop.

This time SHO

  1. Channel combination SII to Red, Ha to Green, OIII to Blue
  2. Screentransfer function looked ok with the default so went straight to Histogram transformation
  3. HDR Multiscale transform - 6 layers, number of iterations 1 with a scaling function of B3 Spline (5) set to lightness mask
  4. Ran the script Dark Structure enhance with the default settings
  5. Then did a starnet to create a star map and a starless image
  6. Inverted the starless image and did a SCNR to remove any green
  7. Did the same on the star map to remove green
  8. Reinverted both images again
  9. Applied 4 iterations of unsharp mask
  10. Recombined the two images with 100% on both
  11. Saved as a 16 bit TIF for import into Photoshop
  12. From Photoshop ran the Topaz Denoise AI plugin
  13. Did a final sharpen on the image
  14. Save as png file

I quite like the structures it has brought out this time.

M17V2TMC.thumb.png.b299db93a142f5b676908f8d8e00f5f8.png

 

Edited by TerryMcK
reduced file size
  • Like 3

Share this post


Link to post
Share on other sites

My second attempt.. SHO in StarTools 1.7436 + Lightroom 

Load all channels saved in mono as luminance, then again in Compose module this luminance and red/sii, green/ha, blue/oii.

Contrast>HDR>Decon>Sharp>Color>Entropy>Color>Shrink>Denoise.

Then further tweaking in Lightroom ^^ 

 

597069626_05112020-Compositem17Procshrinked.thumb.jpg.62d09d8026c2261e8e01d74212e4d143.jpg

  • Like 1

Share this post


Link to post
Share on other sites

Here's a second attempt (sufficiently different, I hope!)

Processed in StarTools 1.7.439

NewComposite3.thumb.jpg.b83f47ae50d8dd62765140b7b2ea7c3b.jpg

  • Like 2

Share this post


Link to post
Share on other sites

This is fantastic! Thank you so much for sharing this amazing data, I had a lot of fun processing it.

My workflow was as follows:

  • Small¬†levels stretch of each channel individually, then ran through StarNet++,
  • Imported the starless images into Photoshop and processed individually;
    • Further levels stretch,
    • Contrast boost in Curves,
    • Blurred the surrounds of the main nebulosity using Gaussian blur and masks,
    • Removed any artefacts left from the star removal using the spot healing brush,
    • Removed noise and increased local contrast using Astronomy Tools and Annies Actions.
  • Combined the edited starless images into SHO palette,
  • Played around with selective colour, hue and saturation to fix the colours to my taste,
  • Adjusted the contrast further with the curves tool,
  • Added the edited Ha starless image as a luminance layer,
  • Created the star layer;
    • Combined the original images with their respective¬†starless images using "difference" to remove all nebulosity and keep just the stars,
    • Combined the stars into SHO palette,
    • Ran dust and scratches noise reduction set at 2 pixels to reduce the stars,
    • Removed the magenta and some of the green using selective colour.
  • Combined the star layer using "screen"

M17_Starless.jpg

M17.jpg

  • Like 11

Share this post


Link to post
Share on other sites

Added a red version for good measure

This is 2 images using same data. Stars processed in SHO, nebula  using Ha for red, SII for green. Both  processed independently in PI then added in blend tool

Tools used for both inc Linear Fit, EZ Softstretch  and HDR, Starnet, SCNR green, Dynamic BackgroundExtraction, HDR wavelets, Curves Hue & Saturation, Advance Sharpen, Blend1598316747_blend6copy.thumb.png.c71add338f782d45552cde8144cd4d73.png

 

slightly redder to bring out the faint nebula

 

blend6_ABE.thumb.png.a8b10397a346a3483b9eba0742d7fab3.png

 

Added tweaked LocalHistogram to emphasise nebula swirls

blend4.thumb.png.6429b82244cd5727bdf55528f0b5fc60.png

 

Edited by billhinge
  • Like 1

Share this post


Link to post
Share on other sites

Another amazing set of data - kept my evenings interesting over the past week. Its an extraordinarily difficult image to balance, huge dynamic range. I found it extremely challenging to get the balance right and also put colour across the image that highlighted the different gaseous regions of M17 and where the Ha, Oii and Sii contributes to the overall false colour image. I used pretty much every tool in the box,

Pi  - initial stretch with histogram and creation of starless versions with Starnet++
PS - massive stretching  of each of the starless images to get a colour into the mix, I found getting any Oiii blue into the central region 'difficult'
Pi - used PixelMath to create a large number of different colour mixes - went around this and above steps many times to get a mix I liked
Pi - deconvolved the Ha frame and used enhance dark structures
PS - Annies Actions to create SHO for stars only
PS to mix Ha as luminance into the Starless colour image
... note I now had white stars as I had used Ha with stars through the de-convolution process, so now had to make the colour image starless again in Pi Starnet++
PS - added in 'colour' stars
PS - reduced stars
PS- selective sharpening and noise reduction using layer masking
PS - very slight crop as merging all the starless images for colour, even wih quite a bit of effort to make background same levels I ended up with a blue edge, so its been cropped
 

So after all that here is my contribution to this growing thread of fantastically different interpretations of M17 from the same data set.


Bryan

 

M17_SHO_Processed_V6.thumb.png.23408cca9a7b6df62a30d40f2b5efd1f.png

  • Like 8

Share this post


Link to post
Share on other sites

HOO version photoshop curves, selective colour, masks and cropped.

 

 

 

M17-compoHOO1.jpg

Edited by Andy R
  • Like 2

Share this post


Link to post
Share on other sites

Fabulous data set, thanks for the opportunity to play with it.

Processed using StarTools 1.7.440a. 
Compose using Ha as green, Sii as red and Oiii as blue, synthetic luminance created from all three.
Cropped 20 pixels off all sides
Bin 50% which also makes processing faster
Wipe as narrowband
Autodevelop to see what we have, ignore fine detail set at 5 pixels
HDR to reveal DSO core
mask created Decon a light touch to reveal the cloud subtle structures
mask created Wavelet Sharpen dark structure enhancement, lightly
mask all Colour set to [SHO 60SII+40Ha,30SII+30Ha+40OIII,100OIII]
mask created Shrink set to tighten
mask reused from Wavelet Sharpen, Super Structure strength just 25% 
mask all Wavelet De-Noise grain size just 4 pixels 
Saved this process

Space to me is three dimensional and not flat but infinite and wanted to reflect this in the process.
I needed a depth map but wanted this to be starless.

Loaded the saved file from above and created a star mask
Heal with new darker than old
Starless image saved
mask all Layer and Undo->Bg and set layer mode to subtract
Stars image saved

Opened the saved starless background image and used the Stereo 3D Synthesis module to create a depth map.
Contrast module to give it a definition boost 50%
Saved the file

Over to The GIMP and opened the saved starless image and duplicated the background layer
The first layer I selected and from Colour / Components / Channel Mixer and turned off blue and green
Next selected the second layer and set the Mode to screen then back to Channel Mixer and this time turned off red.
Saved the image

Over to PaintShropPro and opened the file saved and the depth map file
Selecting just the first red only layer then Effects / Distortion Effect / Displacement Map and selected the depth map and set the intensity to 13%
Saved the file

Over to StarTools and opened the new displaced background file
Layer and loaded the stars from earlier and set to Add
Saved the file

Irfanview was last and opened the new complied file and saved as png for upload here.

For best viewing see if can borrow some 3D glasses or make some DIY style, I just used quality street film wrappers red and blue from my box of bits and pieces.

EDIT: starting with the saved file before the Heal module I reloaded the star mask and removed 9 large stars then completed the remaining steps to PaintshopPro where I created a mask excluding the 9 large stars and immediate surrounding area before applying the Displacement Map.

 

NewComposite 071120 v1.3 healed starless 2 RG D stars.png

Edited by happy-kat
reprocessed selective Heal and displacement map
  • Like 1

Share this post


Link to post
Share on other sites

First time processing narrowband images and thank you IKI and FLO team for the amazing data. Software used: PS and StarNet++.

Workflow:

1. Auto stretch Ha, Oiii, Sii separately in Camera Raw filter. Equalize function from Olly was applied to the three channels and denoising done for Sii and Oiii channels in Topaz Denoise plugin. Lastly, sharpened using Unsharp Mask.

2. Each channel is imported into StarNet++. Starless images combined as SHO in PS. 

3. Gentle stretch to the image after some light denoising. The bright core is masked with the original Sii channel to recover the details.

4. Selective colour and gentle sharpening.

5. Results from Step 1 were combined as SHO as well to obtain the star field using difference blend mode. Star field is then overlaid on top of the image after Step 4 using screen blend mode.

6. Stars made a little smaller using color range tool and minimum under filter tab.

 

FinalAttempt2.jpg

  • Like 2

Share this post


Link to post
Share on other sites

Processed using IRIS with possibly one of the shortest workflows I've ever used:

  • Equalise histograms and apply selective Gaussian filter.
  • Blend RGB (R = ADD_MAX[0.67S, 0.33H, 0.00O OR¬†1.00S]; G = ADD_MAX[0.13S, 0.43H, 0.33O OR 1.00S; B = ADD_MAX[0.00S, 0.17H, 0.67O OR¬†1.00S])
  • Apply luminance¬†( L = H)
  • ASINH colour stretch, adjust saturation and sharpen

The IRIS ADD_MAX is probably equivalent to the PS LIGHTEN and was used to enhance the S regions in the final  image.

m17.thumb.jpg.089eddd6f73bf6868d0dc82bf66a04be.jpg

  • Like 2

Share this post


Link to post
Share on other sites

A huge thank you to FLO & IKO for this, it's been nothing but clouds here so was great to have something to work on.

My goal was to create a classic Hubble palette version and show the huge dynamic range captured in this magnificent data set of M17.

Keeping the green and aqua tones were vital in this image to really emphasise the richness of signal in all three channels.

Pixinsight

Slight Crop to remove some stacking artifacts

DBE (small amount of gradient to remove)

Ha, OIII, SII => stacked as Master Lum

Noise reduction and Deconvolution were not needed as the data is so good

Channels combined with PixelMath in classic Hubble palette

Red: Sii,  Green: Ha, Blue: OIII

 

Several small stretches with Histogram Transformation followed by a Masked stretch

I felt it was very important not to remove all the green as I felt this was a key part of the nebula showing the range of data where Ha and SII were both strong so SCNR was used at a very low level and with particular attention to masking.

LRGB combined with masterLum

Starnet to remove stars and process separately (N.B. lowered stride value to avoid artifacts)

HDR Multiscale Median Transform applied to main nebula

LocalHistogramEqualisation

 

Stars layer

Photometric colour calibration for background neutralisation (narrowband mode)

Curves (saturation)

Star reduction - Morphological Transform

 

Export to photoshop

several iterations of curves using layers, masking, and blending

Adjust colour and saturation to taste

Add stars layer (linear dodge(add)) over the background layer

M17-SHO-V2.thumb.png.09a8be802bcec2ffffb1b5fe3b0d0298.png

  • Like 1

Share this post


Link to post
Share on other sites

I know I had said I was finished with my attempts, but the data is so go that it's just so enjoyable to keep coming back to.  I've also rotated this time to have North pointing up.  I've made with and without stars versions this time, and just for fun there's a GIF looping to show the stars appearing and disappearing - I'm not completely sure if this is within the rules, so apologies if it isn't.1127915454_M17withstars.thumb.jpg.4326bd7e37a1cd3691b43d9fe271cb53.jpgM17.thumb.gif.a7da92f05749c29de6f725ae9a2b8b1f.gif1173642681_M17starless.thumb.jpg.a58dd99b6ab9a5f8433a92763263e2e7.jpg

 

  • Like 1

Share this post


Link to post
Share on other sites

Here's my SHO palette attempt with your data which is very nice and clean indeed. Maybe i need an astrophotography holiday. All processing done in Photoshop.

Thank you for releasing this data and letting us play with it.

M17 SGL.png

  • Like 2

Share this post


Link to post
Share on other sites

My goal was to show the huge dynamic range captured in this magnificent data set of M17 using a ‚Äúnon-standard blend‚ÄĚ to really highlight the main nebula and background nebulosity.
A huge thank you to FLO & IKO for this, it's been nothing but clouds here.  [Processing steps detailed below image.]

M17-V7.thumb.png.f94f6cb529c0fefa562bfc5c6a5d35c8.png

Pixinsight
Slight Crop to remove some stacking artifacts
DBE (only a small amount of gradient to remove)
Ha, OIII, SII => stacked as MasterLum
Noise reduction and Deconvolution were not used as the data is so good
Channels combined with PixelMath
Red: Sii 80% + Ha 20%
Green: Ha 80% + Oiii 20%
Blue: OIII 100%

Several small stretches with Histogram Transformation followed by a Masked stretch
SCNR (2 x 25% with masking)
LRGB combined with masterLum
Starnet to remove stars and process separately
HDR Multiscale Median Transform applied to main nebula with masking

Stars layer
Photometric colour calibration for background neutralisation (narrowband mode)
Curves (saturation)
Star reduction - Morphological Transform - 2 versions targeting large and small stars lower limit (Chris Woodhouse method)

Export to photoshop
several iterations of curves using layers, masking and blending
Adjust colour and saturation to taste

Unsharp mask

Add stars layer (linear dodge(add)) over the background layer

Fin.

 

Edited by Rob63
Use .png (compressed) file for better reproduction
  • Like 2

Share this post


Link to post
Share on other sites

HSO in StarTools 1.7440 and finished in Lightroom

Tried to get more natural coloured stars : ) 

08112020-HSO-M17.thumb.jpg.1a2f6e1b27157d0bfe2d111c46283544.jpg

 

  • Like 1

Share this post


Link to post
Share on other sites

Hi,

 Wonderful data, thank you so much..

 Processed this primarily in PixInsight, finished off in Photoshop. Images from the three filters were sretched and then made starless with Starnet. Combined using SHO_AIP script, to a coloured starlesss image. Luminance layer created from the Ha alone. Coloured and luminance layers converted to Tiff format and combined in Photoshop. Bit of colour adjustment and a tiny bit of selective sharpening added.

 

f)M17_SGL_Post.jpg

  • Like 1

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.

√ó
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.