Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

First DSO Attempt M45 - Help with weirdness!

  • Please log in to reply
39 replies to this topic

#26 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 13 January 2018 - 07:37 PM

the Elf

 

You've given me a lot digest. I need to step back a moment and take this all in. No clear skies in sight on the 15 day forecast, so I guess I've got some time.



#27 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 13 January 2018 - 08:49 PM

Could you please upload an integrated stacked image? Thanks!


I also updated the Dropbox link above to include my master calibration frames and one of my integration attempts.

#28 entilza

entilza

    Soyuz

  • *****
  • Posts: 3532
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 14 January 2018 - 12:20 PM

Hey, I didn't get find your integrated image, but I did download your data and took a look.

 

So yes the camera seems to be an issue, I do find it surprising it is directly related to the brighter stars, again you did overexpose quite a bit that may help if you cut that down.

 

In future if you were to dither each frame you may be able to significantly reduce the effects of the camera, however your dithers would have to be very large.

 

One trick I was able to use is the Canon Banding Reduction script in Pixinsight, the trick here is to rotate your image so the banding is horizontal then apply it.  It helped quite a bit, unfortunately did not remove it completely.  I used "Amount 1.2 and Protect from highlights"  (Scripts->Ultilities->CanonBandingReduction)

 

The workflow was mostly just running multiple DBEs in order to deal with the gradients, then using the Canon banding reduction.  I did not do any noise reduction either just focused on colour calibration and that banding.  I did a slight curve on the inverse range mask to reduce the background slightly and then a slight curve boost.   I also did a few clone stamps to deal with the lines but just a bit to show it can help, you could go further with it to finish up the image. 

 

I would definitely try your camera again with must less exposure time and see how that affects the results.  If you find it's still a problem and enjoying the hobby then you can decide on a new camera.

 

Anyway it is a very nice M45 and for a first image I'd say you did a great job.

Attached Thumbnails

  • monkeybird_m45_horiz_1600B.jpg

Edited by entilza, 14 January 2018 - 12:21 PM.

  • the Elf likes this

#29 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 14 January 2018 - 12:52 PM

Hey, I didn't get find your integrated image, but I did download your data and took a look.

 

So yes the camera seems to be an issue, I do find it surprising it is directly related to the brighter stars, again you did overexpose quite a bit that may help if you cut that down.

 

In future if you were to dither each frame you may be able to significantly reduce the effects of the camera, however your dithers would have to be very large.

 

One trick I was able to use is the Canon Banding Reduction script in Pixinsight, the trick here is to rotate your image so the banding is horizontal then apply it.  It helped quite a bit, unfortunately did not remove it completely.  I used "Amount 1.2 and Protect from highlights"  (Scripts->Ultilities->CanonBandingReduction)

 

The workflow was mostly just running multiple DBEs in order to deal with the gradients, then using the Canon banding reduction.  I did not do any noise reduction either just focused on colour calibration and that banding.  I did a slight curve on the inverse range mask to reduce the background slightly and then a slight curve boost.   I also did a few clone stamps to deal with the lines but just a bit to show it can help, you could go further with it to finish up the image. 

 

I would definitely try your camera again with must less exposure time and see how that affects the results.  If you find it's still a problem and enjoying the hobby then you can decide on a new camera.

 

Anyway it is a very nice M45 and for a first image I'd say you did a great job.

Thats a huge improvement overall! I'm impressed with the background extraction you were able to do. That is a bit of a weak point for me. Also, I was wondering why the canon banding script wasn't working for me. I realize now I need to rotate the image! I'll give this another go, and I think if I can get it to the point that you did, and with a little patience, I can do a bit more with the noise reduction and heal some of those lines. 

 

I was able to grab a Honis modified T3i/600D with a Astronomik CLS-CCD clip in, and I'll retry the same target when conditions permit.

 

Do you still have that range mask cued up? I love to see a screen shot or the settings you used.



#30 entilza

entilza

    Soyuz

  • *****
  • Posts: 3532
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 14 January 2018 - 04:20 PM

Wow you got a camera arleady, cool smile.gif  

 

Here is a shot of the background processing, I did an ABE first with a function degree of 1.  Then Multiple DBE's which I think is important.  I did at least 4 passes adjusted the Tolerance only.  around 1-2.  Try experimenting to slowly get it cleaned up.

 

 

 

 

 

Attached Thumbnails

  • monkeybird_m45_dbe.jpg


#31 entilza

entilza

    Soyuz

  • *****
  • Posts: 3532
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 14 January 2018 - 04:21 PM

It then looked like this after all the background work and the Canon banding script.

 

 

Attached Thumbnails

  • monkeybird_m45_dbe_beginpro.jpg


#32 entilza

entilza

    Soyuz

  • *****
  • Posts: 3532
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 14 January 2018 - 04:22 PM

Here's after the Histogram Stretch and then showing the range mask the range mask becomes cleaner once the background is cleaned up as you can see.

 

Here's also the workflow.  At some point I did some cropping as well.

 

Most of the range masks looked similar to that.  I showed you the settings for one I use, but use the preview to see what its going to look like.

Attached Thumbnails

  • monkeybird_m45_rangemask.jpg

Edited by entilza, 14 January 2018 - 04:23 PM.


#33 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 14 January 2018 - 05:02 PM

Good stuff entilza, thanks for posting that. I'm going to give it a go tonight. As for the camera yeah, it was a bit of a snap decision. I justified it by deciding I could grab this well known camera that plugs in to my existing setup, or spend many nights obsessively researching, purchasing, and setting up a new camera system. Didn't make sense at this early stage for me. In the future I would love to have an autofocus system, filter wheel, and mono cooled setup, but it will have to wait for now.

 

I'm starting the processing again fresh as I type. Going to play with the cosmetic correction process a little more. I don't fully understand the cold pixel sigma, and when I apply what I think is a modest amount it says in will be replacing hundreds of thousands of pixels. 

 

Still, I need to get some less saturated/exposed data. But in the mean time this is kind of fun to try and figure out.


Edited by Monkeybird747, 14 January 2018 - 06:42 PM.


#34 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 15 January 2018 - 02:26 AM

This is officially the best I can do with this data. Still lost a lot of the whispy stuff along the way. Probably in one of the noise reduction scripts.

 

I'm putting this one behind me and chocking it up to experience. It turned out better than I had hoped. Some heavy-handed use of spot healing tool in PS as the very last step to get rid of the worst of the black lines. The rest in PI. Made the top one a little brighter. Thanks for everyone's help on this. I'll post a follow up when the new camera gets here, and the "new equipment clouds" clear up.

 

MB

 

M45 Final Integration Attempt Non linear complete2
 
M45 Final Integration Attempt Non linear complete

 


Edited by Monkeybird747, 15 January 2018 - 02:27 AM.


#35 silvertree88

silvertree88

    Lift Off

  • -----
  • Posts: 23
  • Joined: 14 May 2016
  • Loc: Round Rock, TX

Posted 16 January 2018 - 04:52 PM

Did a quick run through with DSS and Photoshop and its a great M45 especially for your first try!

Attached Thumbnails

  • 2dark.jpg


#36 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 16 January 2018 - 06:15 PM

Did a quick run through with DSS and Photoshop and its a great M45 especially for your first try!


Whoa, you did a great job pulling the faint detail out of that background without destroying it. Nice touch.
  • silvertree88 likes this

#37 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 20 January 2018 - 01:49 AM

A long update, but at this point this thread is serving as more of a AP diary for me to reference later. Hopefully someone else can get some useful tidbits as well.

 

So while waiting for the new camera to arrive I decided to take advantage of the clear night and try a different ISO coupled with the exposure recommendations of 25-30% with the 1100D, before sending it into retirement. I also used the opportunity to make a maiden run of SGP Pro. While there was a bit of a learning curve, SGP Pro worked beautifully and I executed my first automated meridian flip, as well as my first plate solve. The plate solving routine using the embedded PlateSolve2 system was very quick and I couldn't be happier (unless I had a camera rotator and auto focus...and...and...and...). 

 

I had a very aggravating, migrating dust mote (or something similar) through the night. On top of that, the images were upside down after the meridian flip. I thought I could just run them through PI calibration as normal, but something was up because the flats created reverse motes on the tops and bottoms of the stacked image. Whatever was in my image train was always on the bottom of the sub, but moved about here and there as the night went on. I had blown off the camera sensor with a bulb, and cleaned my objective the day before. Maybe something was frosting up? I feel like I should have rotated the flats or lights for the images taken after the meridian flip. Do you typically rotate your camera after a flip, or just let it shoot upside down images?

 

Auto-guiding was ok, but I'm still not sure exactly what I'm looking at in terms of graphs and data. The stars are round-ish, but have a hint of an egg shape. Still work to do there. Total RMS was around .50-.70 most of the night. Towards the end it was creeping up, but then I noticed I was guiding through the tree tops for the last few frames wink.gif. I use the iOptron balance routine, but I'm not sure if it balances east heavy. I might try moving a counterweight a smidgen toward the OTA and see what happens. The iEQ45 Pro is so stiff in RA you can't just unlock the clutches and see where its at. Should I also be balancing camera-heavy?

 

Thats the good news. The bad news is that even at 1600 ISO, which is recommended for the 1100D, and even with 60s and 90s exposures with histo 25-30%, I still have some pretty severe banding/black lines below bright stars. Just below though. I used the banding script, but applying enough to get rid of the lines below the stars creates an opposite artifact above the stars. So I pressed on with processing normally and just used liberal amounts of the healing brush in photoshop as the very last step. The image is ok I guess. I chose M42, and have 1.5 hours of integration time with 40 subs at 90s, and 30 or so at 60s. I calibrated each exposure set own its own, and then combined them starting at the debayer step, just before integration. More integration time is needed for those faint wisps. Dark site wouldn't hurt either wink.gif. Stars look a little fuzzy to me. Another good, learning-filled capture and processing experience!

 

Any who, the new 600D Honis full spectrum with CLS-CCD is here, so the 1100D is out. I'm looking forward to trying out the lp filter from my home.

 

P.S. I gotta say, once I figured out the plate solving I was like a giddy teenager just slewing, solving, and snapping test exposures all over the sky, just to see what was out there. It was the most exciting part of the night! I was completely blown away by what lies just outside of our visual capabilities.

 

M42 Integration Post Complete Crop And Heal

Edited by Monkeybird747, 20 January 2018 - 02:23 AM.

  • entilza likes this

#38 the Elf

the Elf

    Viking 1

  • -----
  • Posts: 716
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 20 January 2018 - 08:23 AM

@Martin: excellent processing! I did not expect the Canon banding script could deal with that stuff. Very well done.


  • entilza likes this

#39 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 20 January 2018 - 01:34 PM

@Martin: excellent processing! I did not expect the Canon banding script could deal with that stuff. Very well done.

+1

 

I used some of his techniques in my last workflow and they made significant differences. 



#40 Monkeybird747

Monkeybird747

    Mariner 2

  • *****
  • topic starter
  • Posts: 259
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 20 January 2018 - 01:49 PM

Here is a before shot of the M42 image from above. Calibrated and integrated with localized normalization and drizzle integration, linear only. You can see the major doughnuts from the moving artifact. Actually they are more like eclairs:) They are also symmetrically out of place. Something needed to be flipped during calibration. I think. The flats were shot after the meridian flip, and used to calibrate subs take from both sides of the meridian. I calibrated the darks with common master bias, but with darks that matched each set of sub exposures (60s and 90s). I did not touch the camera after the flip.

 

Screen Shot 2018 01 20 At 13.20.48
 
And since this was my first time dithering during capture I was able to use drizzle integration. Here is a side-by-side comparison of the same area with and without drizzle. Makes a subtle difference. The downside is it probably quadrupled my total processing time, if not more. The image that drizzle integration produces is massive, and makes all subsequent processing steps take a little longer. I think I read it's something to do with upscaling (whatever that is) during drizzle integration.
 
Screen Shot 2018 01 20 At 13.28.27

 

I'm going to start a new thread with my image capture and processing steps from start to finish on the M42 image with hopes that someone could point out some areas for improvement in my workflow. Off to read up on cable management!

 

MB




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics