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

False Colours in Narrowband Imaging with ZWO filters?

  • Please log in to reply
9 replies to this topic

#1 Cmahesh

Cmahesh

    Vostok 1

  • -----
  • topic starter
  • Posts: 114
  • Joined: 23 May 2016
  • Loc: Galle, Srilanka

Posted 17 February 2019 - 06:01 AM

I recently attempted my first narrow-band image (Rosette Nebula) using my ZWO ASI1600MM-C camera along with ZWO 1st generation narrow-band filters. I did an interim processing of the data I acquired so far and the result is disappointing. The image shows false colours. For example, instead of bluish inner zone, mine is purple! I used 20 luminance frames and 30 SHO frames (10 for each) each 3 minutes. I know that I need to have more SHO data but I doubt whether this false colouring is solely due to relatively smaller amount of data or is it due to poor quality of ZWO NB filters. Actually when I see through that H-alpha filter it is more greenish than reddish. Have anyone else encountered similar issues with ZWO 1st generation NB filters?

 

 

Thank you

Attached Thumbnails

  • combine-SHO-image - quick partial process.jpg

  • geethq and ks__observer like this

#2 CharlesW

CharlesW

    Long time member

  • ***--
  • Posts: 3474
  • Joined: 01 Nov 2012
  • Loc: Chula Vista, CA

Posted 17 February 2019 - 11:08 AM

I would hazard to say that this is more a processing issue and not a filter problem. NB filters don’t really have assigned colors so your end result is due to the blending combinations you’ve chosen in PI. I would do a Google search for Pixelmath NB recipes and play with those. 



#3 H-Alfa

H-Alfa

    Mariner 2

  • -----
  • Posts: 242
  • Joined: 21 Sep 2006
  • Loc: Spain

Posted 17 February 2019 - 11:27 AM

Your problem is which channel you assigned to each filter. Search for "hubble palette" and you will find more suitable combinations.

Enviado desde mi MI 5X mediante Tapatalk

#4 Cmahesh

Cmahesh

    Vostok 1

  • -----
  • topic starter
  • Posts: 114
  • Joined: 23 May 2016
  • Loc: Galle, Srilanka

Posted 17 February 2019 - 07:10 PM

Thank you for the replies. Anyway I used the Hubble Pellete combination. That is S for red, H alpha for green and OIII for blue. Additionally I also combined luminance data using LRGB combination. I also tried without luminance but still the result is the same. I will upload the stacked images of each channel soon.



#5 BenKolt

BenKolt

    Surveyor 1

  • -----
  • Posts: 1749
  • Joined: 13 Mar 2013

Posted 17 February 2019 - 10:22 PM

In addition to the SHO formula that you choose, you can also adjust the color balance after creating the RGB image.  This is where you have absolute freedom to (1) choose your general palette and (2) balance the colors within that palette to achieve your desired effect.

 

Please note that your colors aren't "false" at all in that you are merely representing your three NB line data using the method of your choice.  I always bristle at the use of the word "false" in that it sounds like we are trying to pull off a deception.  There's no deception here at all, just making a choice how to represent the data in color!

 

Best Regards,

Ben


  • DSO_Viewer likes this

#6 Jeff2011

Jeff2011

    Gemini

  • *****
  • Posts: 3494
  • Joined: 01 Jan 2013
  • Loc: Sugar Land, TX

Posted 17 February 2019 - 10:56 PM

I would ditch the lum data and get more narrowband.  To get a more RGB like color image, try the HOO pallet.  You can also blend the Sii with the Ha in the red channel and small amounts in the green to get more of the Oiii data to show up as blue. There is a huge amount of blending variations to try.

 

Another alternative is to collect RGB data and enhance it with narrowband data. 



#7 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 22919
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 18 February 2019 - 12:38 PM

I recently attempted my first narrow-band image (Rosette Nebula) using my ZWO ASI1600MM-C camera along with ZWO 1st generation narrow-band filters. I did an interim processing of the data I acquired so far and the result is disappointing. The image shows false colours. For example, instead of bluish inner zone, mine is purple! I used 20 luminance frames and 30 SHO frames (10 for each) each 3 minutes. I know that I need to have more SHO data but I doubt whether this false colouring is solely due to relatively smaller amount of data or is it due to poor quality of ZWO NB filters. Actually when I see through that H-alpha filter it is more greenish than reddish. Have anyone else encountered similar issues with ZWO 1st generation NB filters?

 

 

Thank you

Your palette here is actually HSO, not SHO. These acronyms indicate which narrow band channel was assigned to what color channel.

 

SHO means you mapped S (SII) to red, H (Ha) to green and O (OIII) to blue. HSO means you mapped H to red, S to green and O to blue. Your image is definitely HSO mapping, rather than SHO mapping. The "standard Hubble" palette is an adjusted SHO mapping where the green channel curve is pulled down a bit, so if you want that nice "blue/gold" palette, you want to do SHO. Usually, SHO will show up very green at first...you need to pull down the green curve a bit to attenuate that green and balance the colors out.


  • 2ghouls likes this

#8 BenKolt

BenKolt

    Surveyor 1

  • -----
  • Posts: 1749
  • Joined: 13 Mar 2013

Posted 18 February 2019 - 01:29 PM

Thank you for the replies. Anyway I used the Hubble Pellete combination. That is S for red, H alpha for green and OIII for blue. Additionally I also combined luminance data using LRGB combination. I also tried without luminance but still the result is the same. I will upload the stacked images of each channel soon.

Jon is right.  I apologize, Cmahesh, that I didn't notice this before.  It did seem odd to me that your supposed SHO image was dominantly red instead of green.  If you are wishing to further process your HSO image, you can apply the same technique to pull down the red channel in this case and see if you like the results.  Most folks seem to prefer either SHO or HOO, but nothing stops you from exploring other palettes.

 

Ben



#9 pyrasanth

pyrasanth

    Surveyor 1

  • *****
  • Posts: 1542
  • Joined: 08 Jan 2016

Posted 18 February 2019 - 02:32 PM

See this it will help http://bf-astro.com/hubblep.htm



#10 Cmahesh

Cmahesh

    Vostok 1

  • -----
  • topic starter
  • Posts: 114
  • Joined: 23 May 2016
  • Loc: Galle, Srilanka

Posted 24 February 2019 - 03:02 AM

Finally I found the real problem which is embarressing. I have swapped the Ha and SII filters in the filter wheel. Then I mapped the channels in the correct way and used SCNR to get rid of green. This is the result. Anyway amount of data I have is not enough to create a good NB image. Thank you everyone who helped me.

Attached Thumbnails

  • NB Quickprocess 2.jpg



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