Jump to content

sgl_imaging_challenge_2021_2.thumb.jpg.72789c04780d7659f5b63ea05534a956.jpg

IKO M16 - Processing Competition


Recommended Posts

  • Replies 56
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

Thank you to all the team at FLO for funding this incredible project, the data is an absolute treat to work with. Here is my first stab at it, configured as HaSHO. Processed in PixInsight and fin

This is my process in Photoshop.  Mostly levels and curves, selective colour, and selective high pass filter.   Plus a Crop which I flipped and orientated to what I feel is the right way up. 

Great to see the variety of images posted, here's my effort. I've gone for a fairly traditional Hubble palette leaving in a few green hints.  I also wanted to give a bit of emphasis to the bright

Posted Images

15 minutes ago, TerryMcK said:

Just a minor one. Is it me or is the data a mirror image of M16? Not that it matters.

Every picture I've seen in the past has the hook and the pillars of creation the other way around.

Miror mirror on the the wall which is the farest of them all?

Regards Andrew 

  • Haha 1
Link to post
Share on other sites

Heres my rendition , not really used to false colour narrowband ,all part of the learning curve thou.. all done in photoshop, struggled with selective colour but happy with results.. thanks for sharing nice clean data..

image.thumb.png.6131f829b0366316f688149be9332cfe.pngM16 FLO COMP.tif

Edited by newbie alert
upload probs,still not sure if its done
  • Like 6
Link to post
Share on other sites

Great to see the variety of images posted, here's my effort.

I've gone for a fairly traditional Hubble palette leaving in a few green hints.  I also wanted to give a bit of emphasis to the bright central area making the eagle stand out a little.

Processing was a mixture of PI and PS.  I stuck with Ha for the luminence sinceI found that it was brighter in all areas than OIII and SII.  Deconvolved in PI along with a a little HDRMT.   Starnet to create starless lum and RGB.  Everything else was PS.  Happy to provide more info on the work flow if wanted.

I have played around a bit with different versions including one where the aim was to highlight all the dust but in the end I went for this  version since I think the eagle is the star of the show.

 

 

M16_IKI_2.jpg

  • Like 14
Link to post
Share on other sites

I've had another play with the data, this time concentrating on the Pillars of Creation. I also mirrored the image after mapping to RGB  as online versions were like that. Once non-linear I extracted a luminance layer on which I used HDR to bring out the Pillars. Once recombined I used a strong application of Morphological Transformation and then unsharp mask. Checked colour balance then applied unsharp mask once more and checked black point.

All done in Pixinsight. I hope I'm improving slowly.

I hope this is sufficiently different to my first version.

 

Pillars of Creation.png

  • Like 4
Link to post
Share on other sites

This has been great for me as I am trying to master learn Pixinsight and have all the books but still not very good at it. I have often blamed my data for being poor and so my images whilst okay were never outstanding for me. Now I am not saying that statement is not true but I think having known good pre-processed data really helped as I knew what could be achieved from this data and so did not give up. I really want to go back now with some of my better old data and try again with what I have learned.

This is my first attempt using only PI and I know there is more I can bring out of the image but thought I would at least present a image in case I do not get the time to improve my efforts. If I improve it later I will replace this image. It does look very similar to some of the above Hubble Pallet entries which gives me encouragement my processing is going in the right direction (although maybe not for winning the competition 😞 ).

Anyway ramble over here it is.

First_Attempt.thumb.jpg.2020a7cc770addf8b8515c5c6028b6b2.jpg 

Steve

Edited by teoria_del_big_bang
  • Like 10
Link to post
Share on other sites

I just couldn't resist having a play with this wonderful data to try creating something different.
I first created the RGB image in Photoshop.
Ha - Red. OIII - Green. Si - Blue.
All subsequent image processing was performed using Affinity Photo.

Duplicated layer to create three identical layers. 
Used curves to tweak out the fainter material in each layer to varying degrees.
Strong curve adjustment on bottom layer. Medium curve on middle. Very subtle curve on upper. 
Created and applied layer masks on the upper two layers created from the layer below it.
Flattened the image.
Duplicated layer and made the upper layer a luminosity layer.
Made saturation adjustments on the lower layer.
Tweaked the individual RGB colour curves to adjust the colour to try and get something a little bit different.
Applied a subtle Live Layer Unsharp Masking adjustment to the upper luminosity layer.
Then applied the Topaz AI Denoise Plugin.
Made a few more subtle curves and levels adjustments on the luminosity layer before flattening, cropping, flipping and rotating the image.

Eagle-RGB-P-Ph-Crop.png

  • Like 7
Link to post
Share on other sites
1 hour ago, Dave said:

I just couldn't resist having a play with this wonderful data to try creating something different.
I first created the RGB image in Photoshop.
Ha - Red. OIII - Green. Si - Blue.
All subsequent image processing was performed using Affinity Photo.

Duplicated layer to create three identical layers. 
Used curves to tweak out the fainter material in each layer to varying degrees.
Strong curve adjustment on bottom layer. Medium curve on middle. Very subtle curve on upper. 
Created and applied layer masks on the upper two layers created from the layer below it.
Flattened the image.
Duplicated layer and made the upper layer a luminosity layer.
Made saturation adjustments on the lower layer.
Tweaked the individual RGB colour curves to adjust the colour to try and get something a little bit different.
Applied a subtle Live Layer Unsharp Masking adjustment to the upper luminosity layer.
Then applied the Topaz AI Denoise Plugin.
Made a few more subtle curves and levels adjustments on the luminosity layer before flattening, cropping, flipping and rotating the image.

Eagle-RGB-P-Ph-Crop.png

Very nice, there's almost a 3D effect around the Pillars of Creation.

  • Like 3
Link to post
Share on other sites

Had another go with more time on it. Tried hard not to go overboard on the sharpening this time and focus more on the core

If you zoom in on the core, you can tell a noticeable difference, although it's less apparent zoomed out 

SHO.png

Edited by matt_baker
  • Like 5
Link to post
Share on other sites

Here's my attempt using PixInsight.

Basic SHO combination.

Curves to shift the tones towards the Hubble Palette.

SCNR to tone back the greens a little and to suppress the magenta stars (inverted image).

Ha as Luminance, stretched with HT and stars blended from a Masked Stretched version.

HDRMT and LHE to bring out the details in the background and foreground respectively (with suitable masks).

Combine Lum with SHO.

Image rotated and flipped to give conventional orientation of M16.

Cheers

Bill

 

M16.jpg

  • Like 7
Link to post
Share on other sites

Some incredible renditions above, bravo to everyone who has participated in the challenge so far. 

As a final offering I present two mono versions of this image, based on the Hα stack, both with and without stars.

M16_final_2.thumb.jpg.b7efab8395a0278eb1b21c938bef35cc.jpg

 

M16_Final_2_s.thumb.jpg.e22f0b9e41306088701138b760d5239b.jpg

Good luck to everyone!

Cheers

Edited by Spongey
  • Like 5
Link to post
Share on other sites

My humble offering.

I had to run  each data image in FITS format through Deep Sky Stacker to convert it to TIF so that I could open it in Photoshop.  This is 'Hubbled' HOO, with Selective Colour adjustments to bring out the gold hues.

It was a pleasure to work with such good data, and I can't wait for the next one, as well as being determined to improve my own data capture (with, as it happens, my new identical Stellamira ED104)

StevieO

M16 HOO v1.jpg

  • Like 4
Link to post
Share on other sites

My thanks too to FLO for the IKI Observatory project and for making this data available.  As others have said, the data is a pleasure to work with and I have really enjoyed processing it.  I'm loving the various different versions posted thus far.  I've done a number of renditions - pretty much starting from the data stretch each time - and I am posting my favourite six five - three full frame in this post and three two cropped that I will post separately.  I hope that they are sufficiently different not to fall foul of the letter or the spirit of the rules.

The data was deconvolved in PixInsight using the excellent PSFimage script to generate the point spread function (it's sooo quick and easy), a star mask on the brighter stars, and the following settings (first settings for the Ha and SII, second for the OIII):

1873558418_Deconsettings.thumb.png.d08c69b33c5941d63cbc9a7dad819f9b.png

To be honest it didn't make a huge difference but did tighten up the data just a bit.  I then used a very gentle application of MLT noise reduction before moving to Photoshop.  I did the colour combine in PS and then used various sharpening approaches (high pass filter, smart sharpen, unsharp mask) selectively.  The various versions come from different stretches of each channel and different applications of selective colour.

First version is a pretty straightforward SHO palette (slightly updated from the original post):

990120049_SHOfullframe.thumb.png.6db3ae4e75c12d5831942283d91c3c83.png

 

Then this one is with a very different colour mapping but to my eye this highlights a different set of structures and gives pretty good star colour:

1161939977_Purplesfullframe.thumb.png.6e020bcb5fcdc031b1b9bc9fa8c34812.png

 

Finally, here is a rather indulgent homage to the cover of Brothers in Arms by Dire Straits (an album that was a favourite back in the day).  I found I was able to get some pink into the image to contrast with the blue...  I also left some sharpening in the extended gas clouds that I'd usually think was a bit too much but liked in this version (in retrospect I felt that the original was a tad over-sharpened so this too has been tweaked slightly).

1094699775_DireStraitsfullframe.thumb.png.89ff1159d10b48f38471c0c0f74faccc.png

Thanks again for making this data available.

Ian

Edited by x6gas
tweaked the SHO and 'Dire Straits' images slightly...
  • Like 5
Link to post
Share on other sites

Following on from my previous post, here are a couple of cropped versions that, I hope, show off the Pillars of Creation a bit better.  Just shows how well the data holds up in my opinion.  Again the data was deconvolved in PixInsight with a bit of MLT noise reduction but the leg work done in PS.

Here it is in SHO (Note I have updated this to a different version in which I have deliberately left a little green; I know that's not fashionable but I think it adds a little definition):

1856723734_v21-1cropped.thumb.png.1efaa9b2491e9c8d9828d84045991e5d.png

And here is a completely different starless process (which admittedly followed very similar steps) that used Starnet++ in PI.  I really like this one too, but in fairness I don't think that the starless images show off just how good the data is... I can (and have) stare at this one for hours, though.

1848467140_Starlesscrop.thumb.png.d4684de6d3ee90903f55b84e0104c62d.png

Thanks again FLO and IKI obsy team.  I've been having a really tough, stressful, and busy time at work these past couple of weeks and I have really enjoyed playing with this data and it has really helped to keep me sane - particularly since the weather has been so poor that I've not been able to capture any data of my own.  :icon_salut:

Clear skies and stay safe, Ian 

 

Edited by x6gas
updated SHO image
  • Like 6
Link to post
Share on other sites

Here's my entry to the competition. It's the first time I've processed an image in SHO (so thank you for very much for the data, if anything it has been great to practice on). It took me a few attempts to get something I was happy with and I've finally come up with this.

I struggled with the stars on my initial attempts as they were coming out magenta as I have heard is common with SHO, and discovered Starnet++ which was a great help so after initial stretching of the Sii, Ha & Oiii created a starless image for each of the channels before combining into a SHO image. Some further processing (levels, curves, colour balance, adding Ha as luminance layer etc) then added the stars back in by creating another image using just the Ha and Oiii data (HOO), selected the stars and pasted onto this image.

All processing done in Photoshop.

1098837485_SHO020920.thumb.jpg.4d882e5cb7660ec57de4db917276d1fc.jpg

 

Adam

 

 

  • Like 6
Link to post
Share on other sites

I've been really enjoying seeing all the submissions so far - it's a real pleasure to see how much fun you're having with the data and it's very inspirational to see all the different 'takes' on how to process this - each one brings something different to the table!

This is going to be hard to judge!!!

  • Like 4
  • Thanks 1
Link to post
Share on other sites
3 hours ago, Grant said:

I've been really enjoying seeing all the submissions so far - it's a real pleasure to see how much fun you're having with the data and it's very inspirational to see all the different 'takes' on how to process this - each one brings something different to the table!

This is going to be hard to judge!!!

Yes, there certainly are a lot of different ways to skin a NB cat, or should that be an eagle?

  • Like 2
Link to post
Share on other sites

I had one last go at this data and wasn't able to get a better result than my previous, especially in the core which is what I wanted to focus on. 

This challenge has really pushed me and this was the first time processing in SHO too. Thank you FLO and IKO for making this possible and working with such good data.

One thing I would've loved was to get a drizzled integration so you can push the resolution and limits in the core to make it nice and crispy and defined.

Here's a rotated version since I like it more and a cropped, zoomed in one

Good luck everyone!

SHO_rotated.png

SHO_cropped.png

  • Like 5
Link to post
Share on other sites

You'll need to rummage in the cupboard to rustle up some 3d glasses, or in my case a root around in my miscellaneous box where i keep all sorts and fished our a red and a blue coloured quality street wrapper. 

This was quite fun but a challenge too.

First pass process was through StarTools using the Compose module to load the three files, I choose HOO with Ha at 1.35 of total exposure, Sii as 1.31 and Oii as 1 after working out what each was as a percentage of Oii which I set as the reference of being 1.

I would normally Bin an image but I didn't on checking the file, then used Autodev and increased the ignore fine detail. Wipe was next then HDR and set to reveal core, next Contrast though I like to lower it slightly to 50%, off to Decon and selected to mask, I adjusted the setting until I got the effect I wanted but one that didn't cause ringing on the stars. Off to the Colour module selecting an option that I liked the colour blend off and slightly increasing the saturation. With Tracking off I ran denoise and saved. I tweaked it and ran a slight Contrast boost over it.

The image didn't have anything different about it and didn't feel animated, StarTools has a new 3D capability but I didn't want someone else doing the work for me, but I did use StarTools to create a depth map which I saved.

First try and creating the 3d image in PaintShopPro left too much in the foreground. Back in StarTools using the Heal module I first created a star mask and left out the main stars I wanted to keep for the 3d effect, using the mask I then healed out the remaining stars and saved a starless image and a star only image, as well as saving the mask.

Back in PaintShopPro I opened the depth map and my new background and selected the all the image and created a new layer. Both layers were set to blend mode screen. One layer I turned off red using Colour Channel Mixer and on the other layer turned off green and blue. I selected just the blue green layer and now used distortion Effect Displacement Map and loaded the depth map and set the distortion to 13%. Once done I turned the floating selection into a layer. This image I saved as a flattened PNG and loaded back in StarTools and RGB non linear and did not reverse the stretch and using the Layer module I opened the star image and selected Add and set the blend amount to 30%. This was saved as PNG and is the image attached.

328799620_M16NewCompositev1_4v1.1background.thumb.png.22ca175d053a633b6f09cc75861bceac.png

EDIT: looking at log files I loaded Sii as red, Ha as green and Oii as blue, colour was [SHO 60SII+40Ha,30SII+30Ha+40OIII,100OIII] and contrast was lower at 25% but ran it again later at 10%. 

Edited by happy-kat
  • Like 5
  • Confused 1
Link to post
Share on other sites
  • Grant locked this topic
Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.