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

learning NB mono imaging

  • Please log in to reply
11 replies to this topic

#1 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,027
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 23 January 2020 - 07:06 PM

I had some clear skies last night and imaged Seagull

 

HA 180 x 18

Oii   180  x 17

SII "  180 x 17

 

 

I fixed my image train and the guiding, calibration correction was superb. 

 

Created three master lighs, DBE.

 

Now the interesting part for a newb, creating  color out of NB images. I read where the hubble pallet assigns HA is green, SI is red and Oii is whatever color is left over. smile.gif  

Using [L] RGB combo, it was all green, and DBE would not change that. So I just dis some random combination, and it looks like all the stars are green. I don't know, maybe this is what it should be like until it is properly procced out. 

 

here are the original master light frames. 

 

https://www.dropbox....LIGHT.xisf?dl=0

 

https://www.dropbox....LIGHT.xisf?dl=0

 

https://www.dropbox....LIGHT.xisf?dl=0

 

 

that amount of time with my triad and OSC would have made a really nicer image, but i am sure that is because I do not know how to proc NB yet.

Attached Thumbnails

  • learning nb.jpg

Edited by Ballyhoo, 23 January 2020 - 07:08 PM.


#2 mistateo

mistateo

    Apollo

  • *****
  • Posts: 1,310
  • Joined: 02 Feb 2017
  • Loc: San Diego, CA

Posted 23 January 2020 - 08:01 PM

Hey Nick.  Welcome to Narrowband!

 

I gave your data a quick 5 minute process (the most basic narrowband SHO process I could do):

Dynamic Crop: to remove the dark areas

Channel Combination: Sii - R, Ha - G, Oiii - B

Histogram Transformation: Stretch to non linear state

SCNR: Remove green dominance (all default settings)
Invert Image, Apply SCNR, Invert image again: Subdue Magenta Halos around stars

Curves Transformation: Gentle color saturation boost

 

Resample to lower resolution and convert to jpg: for posting here

 

This is not my normal narrowband workflow. There was no noise reduction, sharpening, color calibration using linear fit.  Plus I normally create color image using pixel math to my personal taste rather than Channel Combination.  I just wanted to get a peek at what your data looked like.  It seems like something went wrong during flat frame calibration for Sii channel, as evident in the top right corner especially.

 

I would recommend looking at some narrowband workflow tutorials and following along until you get the hang of what works for you.  My workflow listed above is a quick and dirty example, not advice wink.gif

 

NickEagle.jpg

 

 


  • Ballyhoo, DSOs4Me, Monkeybird747 and 1 other like this

#3 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,027
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 23 January 2020 - 08:20 PM

Thank you Bro!

 

Quite a few members have given me various flows,  but I think much has been LRGB, and still, I just need to put it all together and figure it out.

 

For example, the whole Keller book, when he writes a chapter or anything, is it about OSC, mono, LRGB, NB...

 

When I was in college I got marked down  on a paper I was writing in international politics about N Ireland. I got marked down because much of the time I did not make distinctions bw north and south.      I do not feel that is being done in the Keller book.  There is so much in the Keller book but, I am having issues wrapping my head around differences between yea, NB, LRGB, and OSC proccing in that great book.   Not Keller's fault. I just need to get oriented with this new genre of mono imaging. 

 

Thank you for a great image. 



#4 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,027
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 23 January 2020 - 08:21 PM

Also, do you see how I got sucked down the rabbit hole since that one day I came and grabbed that AVX?? shocked.gif

 

Edit

 

I think someone could dedicate an entire book, just to mono imaging. There are so many possibility, but all the literature convolutes LRGB, OSC, NB and mono imaging together.


Edited by Ballyhoo, 23 January 2020 - 08:24 PM.


#5 mistateo

mistateo

    Apollo

  • *****
  • Posts: 1,310
  • Joined: 02 Feb 2017
  • Loc: San Diego, CA

Posted 23 January 2020 - 08:43 PM

Also, do you see how I got sucked down the rabbit hole since that one day I came and grabbed that AVX?? shocked.gif

Yup.  Now with a mono camera and narrowband filters.  You are one step away from selling organs for optical perfection.


  • Ballyhoo likes this

#6 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,027
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 23 January 2020 - 08:55 PM

thank you, followed your steps, plus added an MLT.

but instead I used color saturation which is likely a bit much. 

 

I am going to add more data.

 

you were correct, and good eye: What happened last night was somehow I ran the SGP calibration wizard and I stopped it prematurely before it completed the Siii flats.

 

I hope I will have organs left to sell. 

 

Attached Thumbnails

  • SeaGull First NB ss.jpg

  • twidget likes this

#7 mistateo

mistateo

    Apollo

  • *****
  • Posts: 1,310
  • Joined: 02 Feb 2017
  • Loc: San Diego, CA

Posted 23 January 2020 - 09:24 PM

thank you, followed your steps, plus added an MLT.

but instead I used color saturation which is likely a bit much. 

 

I am going to add more data.

 

you were correct, and good eye: What happened last night was somehow I ran the SGP calibration wizard and I stopped it prematurely before it completed the Siii flats.

 

I hope I will have organs left to sell. 

Looks like you followed MOST of the steps. One important step you didn't do is crop out the very dark borders on the right and bottom.  This is important because DBE (and other functions) won't work properly with these abnormally dark areas present in the image.  With either crop or dynamic crop (like many other processes) you can set it up once, and apply to all 3 channels.


  • Ballyhoo and Swordfishy like this

#8 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,027
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 23 January 2020 - 09:41 PM

Looks like you followed MOST of the steps. One important step you didn't do is crop out the very dark borders on the right and bottom.  This is important because DBE (and other functions) won't work properly with these abnormally dark areas present in the image.  With either crop or dynamic crop (like many other processes) you can set it up once, and apply to all 3 channels.

great point. 


  • twidget likes this

#9 Madratter

Madratter

    Voyager 1

  • *****
  • Posts: 11,728
  • Joined: 14 Jan 2013

Posted 24 January 2020 - 12:01 PM

There are some simplified processing methods for NB that are out there. And they actually work on some targets. But in general, my opinion is that NB processing is quite a bit more difficult than LRGB. Just make up your mind that it will take some learning and push through it.


  • Ballyhoo and Swordfishy like this

#10 Monkeybird747

Monkeybird747

    Viking 1

  • *****
  • Posts: 720
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 24 January 2020 - 07:29 PM

You'll start to notice that the signal strength in each channel varies from target to target. Ha is often the dominant channel compared to Sii and Oiii. When you do a straight SHO hubble blend the image will often look very green with magenta stars, because you just assigned the most dominant channel to green. DBE is not the tool to fix that. DBE helps you eliminate sky gradients and uneven illumination, but it's not to get the green out of a hubble image. You would need to focus more on color manipulation process, or change the blend in pixel math. The initial combination is just a starting point. Don't expect to get your final color balance with just the pixelmath blend alone.

 

There is a bit of a sledgehammer approach when it comes to reducing the green out of a sho blend. Hit it with SCNR green AFTER you stretch the image. Vary the SCNR amount to taste. To tackle magenta stars with the same hammer, invert the image and the magenta stars will now look green. Hit it with scnr green again, then invert the image back. No more magenta stars.

 

The color mask script is very helpful for narrowband. Instead of the SCNR approach you can create a mask of just the green in the image, then use the mask and curves to manipulate the green into something else.

 

You can also vary the blend for a starting point. Search for narrowband pixelmath blends. Lots out there. If you're looking for more of an RGB-ish look try out some HOO blends. Don't expect to get RGB stars and color palettes with narrowband, unless you add RGB star data later. Just take what the data gives you.

 

You will need to vary the total imaging time for each channel depending on the target. A heavy Ha target with weak Oiii might only need a few hours of Ha to look nice, but you made need triple or more that time in Oiii and/or Sii. Ha always integrates smooth with a pleasing noise profile. Sii always looks grainy to me, even with lots of integration time. Oiii has typically been the weakest for the targets I've shot (except for veil complex). A typical exposure time distribution for me would be something like 5-7 hours Ha, 10 or so sii, and probalbly 15-20 Oiii. Bottom line is a 1:1:1 ratio is probably not going to work for most targets.

 

I know you're still in the proof of concept phase of mono, but you do a lot of one-night images. You will probably need to commit to multiple nights on one target for the results you're looking for. Try shooting just one filter per night, for 3-5 nights. This will save you focusing since you won't be changing filters. Plan your Oiii around new moon. The results will likely blow your mind.

 

Here is your challenge: Pick a target that you can shoot all night that has all three channels present. Shoot just Ha on night one. Shoot just Sii on night two, shoot Oiii for nights three and four. Then post up your 30 hour integration. You'll have an image with amazing depth and smooth noise profile. You'll actually be able to zoom in and still like it.


Edited by Monkeybird747, 24 January 2020 - 07:31 PM.

  • Ballyhoo likes this

#11 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,027
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 24 January 2020 - 08:38 PM

You'll start to notice that the signal strength in each channel varies from target to target. Ha is often the dominant channel compared to Sii and Oiii. When you do a straight SHO hubble blend the image will often look very green with magenta stars, because you just assigned the most dominant channel to green. DBE is not the tool to fix that. DBE helps you eliminate sky gradients and uneven illumination, but it's not to get the green out of a hubble image. You would need to focus more on color manipulation process, or change the blend in pixel math. The initial combination is just a starting point. Don't expect to get your final color balance with just the pixelmath blend alone.

 

There is a bit of a sledgehammer approach when it comes to reducing the green out of a sho blend. Hit it with SCNR green AFTER you stretch the image. Vary the SCNR amount to taste. To tackle magenta stars with the same hammer, invert the image and the magenta stars will now look green. Hit it with scnr green again, then invert the image back. No more magenta stars.

 

The color mask script is very helpful for narrowband. Instead of the SCNR approach you can create a mask of just the green in the image, then use the mask and curves to manipulate the green into something else.

 

You can also vary the blend for a starting point. Search for narrowband pixelmath blends. Lots out there. If you're looking for more of an RGB-ish look try out some HOO blends. Don't expect to get RGB stars and color palettes with narrowband, unless you add RGB star data later. Just take what the data gives you.

 

You will need to vary the total imaging time for each channel depending on the target. A heavy Ha target with weak Oiii might only need a few hours of Ha to look nice, but you made need triple or more that time in Oiii and/or Sii. Ha always integrates smooth with a pleasing noise profile. Sii always looks grainy to me, even with lots of integration time. Oiii has typically been the weakest for the targets I've shot (except for veil complex). A typical exposure time distribution for me would be something like 5-7 hours Ha, 10 or so sii, and probalbly 15-20 Oiii. Bottom line is a 1:1:1 ratio is probably not going to work for most targets.

 

I know you're still in the proof of concept phase of mono, but you do a lot of one-night images. You will probably need to commit to multiple nights on one target for the results you're looking for. Try shooting just one filter per night, for 3-5 nights. This will save you focusing since you won't be changing filters. Plan your Oiii around new moon. The results will likely blow your mind.

 

Here is your challenge: Pick a target that you can shoot all night that has all three channels present. Shoot just Ha on night one. Shoot just Sii on night two, shoot Oiii for nights three and four. Then post up your 30 hour integration. You'll have an image with amazing depth and smooth noise profile. You'll actually be able to zoom in and still like it.

Well,

 

there are a lot of concepts I am not even familiar with yet. I still do not really know how to use PM.

I did join up W Adam Block's fundamentals, but I would expect that will take time to go through them sequentially. 

 

Actually, I have continued to pile on more data to Seagull. I almost doubles the amount of data by imaging last night. So I have the data.  I will use what I have, maybe get more data, to learn this. Because right now I am only getting a couple of tones out of it.

 

That is even though I added data the image has  not yet improved. but I can make it a project to improve it w processing. 

 

It is supposed to be clear tonight /tomorrow. I am going to get some LRGB on this.

 

Not that I will know how to proc it right off the bat, but if the data is good. I will always have it.

 

I also received my Rigel motofocus. 

 

It is not yet installed and I think Leon needs to send me a custom part for it.

Attached Thumbnails

  • SeaGull First NB n2.jpg

Edited by Ballyhoo, 24 January 2020 - 08:39 PM.

  • Madratter likes this

#12 Madratter

Madratter

    Voyager 1

  • *****
  • Posts: 11,728
  • Joined: 14 Jan 2013

Posted 24 January 2020 - 08:47 PM

That is a fairly nice if unconventional looking result. Of course one of the nice things about narrowband is there is some freedom to play with color interpretation.


  • Ballyhoo likes this


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