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

Help with Test of modded Canon dSLR on Cygnus region H-alpha

DSLR Filters Imaging
  • Please log in to reply
10 replies to this topic

#1 IR656nm

IR656nm

    Explorer 1

  • -----
  • topic starter
  • Posts: 87
  • Joined: 12 Mar 2018

Posted 25 October 2021 - 04:55 PM

Hi CloudyNights,

 

I would be most grateful for some help. I image from the suburbs of Greater London, with a Bortle zone of 7, sky quality estimate of 18.41.

I should be doing some narrowband imaging with a telescope, small refractors and a 150mm Newtonian reflector later when I have got my setups correct and my coma correctors, and collimation all correct....

 

I just acquired a modded Canon 100D (from JWT cameras, Holland) earlier this year. I have got a Svbony 7nm H-alpha 2 inch filter. I mount this on top of my Canon 50mm lens (I will get adapters so that I can use it for longer focal length lenses).

I am just playing with the Canon and h-alpha at the moment, trying to learn more about h-alpha channels and image processing and combination.

 

I have got about 25 images of the Cygnus region near Deneb tonight. I took 90 sec exposures at ISO 800 and then took bias and dark frames (no flat frames yet).

Unfortunately, some of the latter frames there was dew on the H-alpha filter.

 

I get a very red image like this image 1; but the red channel looks like this (not much processing here).

Am I supposed to extract the Red channel from the Raw file of each sub, and then stack the result or do I just stack the whole unprocessed RAW file and then extract the Red channel from the stacked result ??

 Can I just  process the Red channel of the Raw file ?? What program would be best for this. I have DSS, no dedicated astro imaging software like PixInsight or APP, or Nebulosity.

 

I am not sure if my subs is any good. I read some posts saying it is a waste of time to image with a modded dSLR and narrowband filters.

Have I wasted so much money then ??

 

Imaging setup: Vixen Polarie tracker polar aligned. Canon 50mm F1.4 lens - at F2.5 ISO800, 90 sec exposures. H-alpha 2 inch filter (Svbony), 7nm bandwidth.

 

Thanks a lot.

 

Magnus.

 

Image 1 (jpg of original, saved as low quality jpeg file) _MG_3808-smaller.jpg

Image 2 (red channel of RAW file, resized down to 30percent)  Cygnus-Red-channel-levels-30pc.jpg


Edited by IR656nm, 25 October 2021 - 04:57 PM.


#2 galacticinsomnia

galacticinsomnia

    Surveyor 1

  • *****
  • Posts: 1,728
  • Joined: 14 Aug 2020
  • Loc: Pacific Northwest - Oregon

Posted 25 October 2021 - 05:36 PM

When it comes to post work, I have to say that Google is your friend here.
Plenty of videos and instructions for various work flows.

Here is the top one on a google search.
https://astrobackyar...raphy-tutorial/

 

Clear Skies !!


  • IR656nm likes this

#3 asanmax

asanmax

    Vendor - DSLR Modifications

  • *****
  • Vendors
  • Posts: 1,344
  • Joined: 17 Sep 2018
  • Loc: Vancouver BC

Posted 25 October 2021 - 11:07 PM

You never waste your time when you're outside imaging. You either enjoy the smooth ride or your learn. 

 

As for using an Ha filter with a modded DSLR, the math is simple. Your sensor can only catch 25% of red photons at a time. If your sensor has the quantum efficiency of around 40%, that means that your camera only catches around 10% of that H-alpha light that comes through the lens. Efficient? NO.

 

And that's why it's a good idea to use multi-band filters when imaging with a modified DSLR or an OSC camera. You can catch Ha, OIII, Hb and SII all at the same time.

 

Red cast in the images that come from a modified camera is totally fine. You just need to color correct the image to make the background sky neutral. Something like R20, G20, B20.


  • Rickycardo and Benschop like this

#4 michael8554

michael8554

    Surveyor 1

  • -----
  • Posts: 1,527
  • Joined: 19 Aug 2020
  • Loc: Wiltshire UK

Posted 26 October 2021 - 04:13 AM

I'd suggest much longer exposures, or higher ISO if quality allows.

 

You're imaging with the mostly Red sensors, through a narrowband filter, so very little light is being recorded.



#5 IR656nm

IR656nm

    Explorer 1

  • -----
  • topic starter
  • Posts: 87
  • Joined: 12 Mar 2018

Posted 26 October 2021 - 05:06 AM

You never waste your time when you're outside imaging. You either enjoy the smooth ride or your learn. 

 

As for using an Ha filter with a modded DSLR, the math is simple. Your sensor can only catch 25% of red photons at a time. If your sensor has the quantum efficiency of around 40%, that means that your camera only catches around 10% of that H-alpha light that comes through the lens. Efficient? NO.

 

And that's why it's a good idea to use multi-band filters when imaging with a modified DSLR or an OSC camera. You can catch Ha, OIII, Hb and SII all at the same time.

 

Red cast in the images that come from a modified camera is totally fine. You just need to color correct the image to make the background sky neutral. Something like R20, G20, B20.

Hello Asanmax,

Thanks for your advice.

Do you mean use a single filter which passes Ha, Oiii, Hb and Siii ? I heard about Optolong duo narrowband filters but I don't think I have heard triband filters or quadband filters. Or do you mean get a filter wheel and image with each filter ?

 

What do you mean R20,G20,B20. Is that the value of the Red channel=20 in the levels. Can you please explain ??

 

When I got the original red image, the Blue and Green channel have almost zero data, just a few hot pixels I think. All the data seemed to be in the red channel. So when I got the Red channel data, there was no colour cast.

 

Thank you

Magnus



#6 IR656nm

IR656nm

    Explorer 1

  • -----
  • topic starter
  • Posts: 87
  • Joined: 12 Mar 2018

Posted 26 October 2021 - 05:14 AM

I'd suggest much longer exposures, or higher ISO if quality allows.

 

You're imaging with the mostly Red sensors, through a narrowband filter, so very little light is being recorded.

Hi Michael,

Thanks for your input.

I thought my detail on the Red channel was reasonable. I will compare them to a high resolution H-alpha scan that was done earlier this century. I am reasonably happy as this was my first test run.

I shall do some calculations, and then try and optimise my exposure times/ISO settings...

 

Magnus



#7 IR656nm

IR656nm

    Explorer 1

  • -----
  • topic starter
  • Posts: 87
  • Joined: 12 Mar 2018

Posted 26 October 2021 - 05:44 AM

When it comes to post work, I have to say that Google is your friend here.
Plenty of videos and instructions for various work flows.

Here is the top one on a google search.
https://astrobackyar...raphy-tutorial/

 

Clear Skies !!

Thanks,

That was a good link. I watched the H-alpha processing of the Bubble nebula which is basically what I needed. Thanks, I will try that process and see what I get...

 

Clear Skies !!



#8 DubbelDerp

DubbelDerp

    Surveyor 1

  • *****
  • Posts: 1,958
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 26 October 2021 - 07:35 AM

I think your data is reasonable for a narrowband filter on a modded camera. I'll repeat the above suggestion for longer subs. With my Skyguider Pro and 50mm lens, I could do 5-minute subs unguided. You'll get much more signal in your images if you leave the shutter open much longer to let it soak up all those H-a photons.

 

Nothing wrong with DSS for stacking and then tossing the blue and green channels in your image to get the greyscale H-a. There are better ways with astro-specific software, though. Just as an example, Astro Pixel Processor has a debayer algorithm specifically for extracting the H-a from the data when using a filter with OSC camera. You'll get much cleaner data this way, but to start, stack as normal and keep only the red channel from your stretched image.

 

Any sensor noise will be amplified when taking long exposures through H-a filter, so make sure your calibration frames are as good as they can be. Darks, flats, dark flats/bias. Dither your exposures if you can.


  • IR656nm likes this

#9 IR656nm

IR656nm

    Explorer 1

  • -----
  • topic starter
  • Posts: 87
  • Joined: 12 Mar 2018

Posted 26 October 2021 - 09:03 AM

Hi CloudyNights,

Thanks for your helpful suggestions. Well I managed to stack 20 out of my 25 light frames that is 90secx20=30 minutes (best 80 percent).

ISO 800 Canon 100D, modded for increased H-alpha sensitivity.

Tracked on Vixen Polarie, Bortle skies 7, Sky brightness 18.41 (Clear Outside app).

 

I followed the defaults for narrowband image/dSLR and finally got this. I used the defaults at first for normal imaging, and got strange results. Luckily I persevered. I think this is ok, now need to go a bit longer, will need to practise this with a HEQ5 Pro mount and then go for guiding and then learn more about settings to use in DSS because getting awful or strange results is not easy to take. In fact, failing to get good results from DSS was why I turned to Sequator, and then I lost a lot of time.

 

Here is a .jpg version, red channel of stacked image was copied into a new file and a 16 bit mono of that image was saved as a 16 bit tiff and also as a .jpg.

 

Thanks for looking.

Magnus.



#10 IR656nm

IR656nm

    Explorer 1

  • -----
  • topic starter
  • Posts: 87
  • Joined: 12 Mar 2018

Posted 26 October 2021 - 09:08 AM

Sorry I forgot to attach it. This is a medium quality .jpg file from the 16 bit tiff (original 32 bit tiff file from DSS stacking).

 

C&C welcome.

 

Merged-H-alpha-mono-levels-curves-16bit-medjpg.jpg

 

Many thanks.

Magnus


  • DubbelDerp and asanmax like this

#11 asanmax

asanmax

    Vendor - DSLR Modifications

  • *****
  • Vendors
  • Posts: 1,344
  • Joined: 17 Sep 2018
  • Loc: Vancouver BC

Posted 26 October 2021 - 11:37 AM

Hello Asanmax,

Thanks for your advice.

Do you mean use a single filter which passes Ha, Oiii, Hb and Siii ? I heard about Optolong duo narrowband filters but I don't think I have heard triband filters or quadband filters. Or do you mean get a filter wheel and image with each filter ?

 

What do you mean R20,G20,B20. Is that the value of the Red channel=20 in the levels. Can you please explain ??

 

When I got the original red image, the Blue and Green channel have almost zero data, just a few hot pixels I think. All the data seemed to be in the red channel. So when I got the Red channel data, there was no colour cast.

 

Thank you

Magnus

Hi Magnus,

 

Yes, the L-eNhance is a tri-band filter (OIII, Hb, Ha), L-eXtreme is duo-band (OIII, Ha). There are also quad-band filters available.

 

If you want to use only your Ha filter, I would, as others, suggest doing longer exposures if you can. 5 to 10 minutes is what typically works great.

ISO 800 is fine with your camera. You can use ISO 1600 if it's cold outside.

 

R20, G20, B20 - this is typically considered the 'normal' sky colour in the astro images. Not necessarily the rule but when the sky doesn't look natural when it's too dark.

The RGB colours are graded from 0 to 255. So you can adjust the image colour balance so that the background sky colour is close to Red20, Green20 and Blue20. It is up to you whether you want it darker or lighter.

This is applicable to full colour images.

 

In your case, you only work with one channel - Red. So essentially you don't need to do any color balancing.

 

Max


  • IR656nm 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





Also tagged with one or more of these keywords: DSLR, Filters, Imaging



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics