Jump to content

SkySurveyBanner.jpg.21855908fce40597655603b6c9af720d.jpg

M81 & M82 First full PI processing workflow. Help and advice please.


Recommended Posts

Bear with me here there were a lot of firsts for me for this image. First galaxies imaged outside of M31. First full workflow with all pre-processing and processing done in PixInsight. A lot of new processes, pixel math etc. First time analysing FWHM and eccentricity. I'm not unhappy with the outcome(s) but I'm sure there's more to be had from the data as well.

 

Data captured with Esprit 100ED and Zwo ASI1600mm Cool.

L: 100x45s

RGB: 20x120s

Ha: 8x300s

~4hrs integration

Calibrated with Darks, Flats and dark Flats

 

Workflow (followed basic workflow outlined in Mastering PixInsight book):

  • Weighted Batch Preprocessing
  • Crop
  • DBE (on each individual channel)
  • Combine RGB
  • Colour balance and Background Neutralisation
  • Stretch RGB, L and Ha
  • Curves and Colour saturation
  • LRGB combination
  • SCNR to remove green
  • Noise reduction
  • HDR compression
  • Pixel Math to add Ha to Red channel

 

The biggest issues I have had were with Dynamic Background Extraction. There were apparently some crazy gradients (see below). I struggled with it a long time until I found something I could live with but that came at the expense of some background detail I could see was there. I've now since learned that it may be because of the IFN in the area. Can anyone comment if the background model below might show that and if there is anything I can do in DBE about it? I saw similar in all channels. It would be great if I could recover some of the background detail. Maybe I can't get the IFN but I think I could bring out more of Holmberg IX (?) just below M81. I actually have another 4hrs of data from the night before but I left focus the same all night and it drifted quite a bit with temperature changes maybe i could add that too. Also can anyone comment on the FWHM and eccentricity plots. I've read anything under 0.4 eccentricity is considered good but does the distribution of it across my sensor tell me anything bad or good? My FWHM seems to drift from left to right.

171021724_Backgroundmodel.JPG.5cc188348542d3dc8e12e680d9c0e06e.JPG412517604_Calibrationanalysis.thumb.PNG.4f25372b5a5d0284b8a7a0a794f3ac4b.PNG

 

The next biggest issue was adding the Ha data and perhaps I should have added it earlier. I could not get all the stars removed using PixelMath. It has definitely added something to the galaxies, not as much as I had hoped but it's there. Unfortunately it has affected star colour to the extent that I think I prefer the LRGB version. If anyone has any tips on adding the Ha I would appreciate it.

 

Finally, if anyone is really interested you can have the data to play with. I'd be really interested to see what someone with more experience could get from it.

 

Thanks for looking and please let me know what else I could do.

 

 

 

2068167900_BodesandCigarGalaxy.thumb.png.9a3d737803469517199b9fc0aafdbc56.png

 

283919544_BodesandCigarGalaxy_HaLRGB.thumb.png.757f35e7920328c99f9991ba7298da7b.png

  • Like 5
Link to comment
Share on other sites

1 hour ago, Icesheet said:

If anyone has any tips on adding the Ha I would appreciate it.

Very nice image - I prefer the first one - great framing.

I'm no expert but I was advised to add Ha to R using a Lighten blend (in PixelMath that would be max(R,Ha) - you can add a multiplier in front of the Ha if you wish, e.g. max(R,(0.7*Ha)). Once the Ha is added to R you can combine with G and B.

There is also an NBLRGBCombination script but I've never used it so can't comment on how well it works.

I'm sure someone who is an expert will pop up and advise.

Adrian

Edited by Adreneline
typo
  • Like 1
Link to comment
Share on other sites

 

1 hour ago, Adreneline said:

Very nice image - I prefer the first one - great framing.

I'm no expert but I was advised to add Ha to R using a Lighten blend (in PixelMath that would be max(R,Ha) - you can add a multiplier in front of the Ha if you wish, e.g. max(R,(0.7*Ha)). Once the Ha is added to R you can combine with G and B.

There is also an NBLRGBCombination script but I've never used it so can't comment on how well it works.

I'm sure someone who is an expert will pop up and advise.

Adrian

Thanks Adrian! Yes, that’s the one I prefer too. 
 

Thanks. I’ll try that too then and see if I get anywhere. I followed a Light Vortex tutorial for adding the Ha and this seems along the same lines but they didn’t weight it so perhaps that will help. I tried the NBLRGB script and that was a complete mess. Probably done something wrong though.

 

  • Like 1
Link to comment
Share on other sites

1 hour ago, Icesheet said:

I’ll try that too then and see if I get anywhere.

Some people advocate using Screen instead of Lighten - you can do that in PixelMath with 1-(1-R)*(1-Ha). Personally I prefer the result using Lighten but maybe it all depends on the data. Experimentation is the name of the game!

Adrian

  • Like 1
Link to comment
Share on other sites

7 hours ago, Icesheet said:

If anyone has any tips on adding the Ha I would appreciate it.

I tried a variety of methods to add Ha to my M81/M82 image and one successful one was using a mask so it only goes where you want it - so, definitely not the background or the stars.  I'm not sure if any of the scripts mentioned doing this as I haven't used them.

As Adrian mentions, experimenting is the way forward.  Methods will vary with when it's added (linear / non-linear) and where it's added.

7 hours ago, Icesheet said:

The biggest issues I have had were with Dynamic Background Extraction

I don't have a lot of experience with DBE but from reading and experimenting myself, that background might not be ideal.  Typically it's removing a gradient so it should be smooth not lumpy.  In areas that look lumpy, look where samples have been placed, remove and re-run.  Always keep an eye on the background model.

I've quickly thrown ~5 hours of Lum into PI and DBE.  It's not perfect (at all) but an example (simple STF - orig, after DBE, background model):

image.thumb.png.2828a586a0230f63b9fffaf2ab00c2c8.png

BTW, great images and processing.  The IFN is REALLY faint and even Holmberg IX - you've done well to pull the latter out cleanly even though it's faint.  You might find the GAME script can help with galaxies and areas of interest you'd like to work on in isolation.

Edited by geeklee
Added example with same area
  • Like 1
Link to comment
Share on other sites

46 minutes ago, geeklee said:

I tried a variety of methods to add Ha to my M81/M82 image and one successful one was using a mask so it only goes where you want it - so, definitely not the background or the stars.  I'm not sure if any of the scripts mentioned doing this as I haven't used them.

That makes sense and I haven't tried that yet. Will see how I get on with that.

 

48 minutes ago, geeklee said:

I don't have a lot of experience with DBE but from reading and experimenting myself, that background might not be ideal.  Typically it's removing a gradient so it should be smooth not lumpy.  In areas that look lumpy, look where samples have been placed, remove and re-run.  Always keep an eye on the background model.

I've quickly thrown ~5 hours of Lum into PI and DBE.  It's not perfect (at all) but an example (simple STF - orig, after DBE, background model):

image.thumb.png.2828a586a0230f63b9fffaf2ab00c2c8.png

That looks a lot more like what I was expecting. Seems I maybe haven't been careful enough with the samples then. Will go back to this too.

 

51 minutes ago, geeklee said:

BTW, great images and processing.  The IFN is REALLY faint and even Holmberg IX - you've done well to pull the latter out cleanly even though it's faint.  You might find the GAME script can help with galaxies and areas of interest you'd like to work on in isolation.

Thank you! That GAME script looks useful, I will try that when trying to add the Ha to the galaxies.

  • Like 1
Link to comment
Share on other sites

DBE can be a mare sometimes. If your struggling just place 4 markers in the corners and experiment with the tolerance. Another idea could be to find a really good image on Astrobin with a similar FOV, plenty of IFN and clear background. Register the image with yours, and use it to place markers over the background. Make a new instance and then apply to your image. I had to do this on an Orion and Horsehead image I done last year.

  • Like 3
Link to comment
Share on other sites

9 hours ago, david_taurus83 said: Another idea could be to find a really good image on Astrobin with a similar FOV, plenty of IFN and clear background. Register the image with yours, and use it to place markers over the background. Make a new instance and then apply to your image. I had to do this on an Orion and Horsehead image I done last year.

I would never have thought of this, great idea!

 

Some good things to work on. Thanks folks

 

 

Link to comment
Share on other sites

  • 6 months later...

Sorry I know this is an oldish thread but on images with very little actual background, so lots of nebulosity or dust clouds or galaxy is DBE really needed ?
I know in many images it is the single most useful tool there probably is and normally is a must do, but with so little you can actually determine as real background is it worth going through all the pain to add the samples in the right place and then play around with the parameters to smooth it al out ??

I am assuming with this that the images also do not show any real obvious gradients or artifacts as well 🙂 

Steve
 

Link to comment
Share on other sites

3 hours ago, teoria_del_big_bang said:

on images with very little actual background, so lots of nebulosity or dust clouds or galaxy is DBE really needed ?

Definitely. One way to place samples is to find an image with the same target online. Crop this image to get approximately the same fov as the image you want to correct. Then use this image to place samples. If you can successfully remove a gradient you will be able to extract more detail where that gradient was most obvious. Sometimes you can only appreciate the difference once you've done dbe.

  • Thanks 1
Link to comment
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.
×
×
  • 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.