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

Veil Nebula with Vintage Lens

  • Please log in to reply
12 replies to this topic

#1 DubbelDerp

DubbelDerp

    Messenger

  • *****
  • topic starter
  • Posts: 422
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 05 December 2019 - 11:01 PM

Well this was an image that just about drove me mad. As a complete AP beginner, with a freshly acquired camera tracker, I set out to capture the veil nebula complex at 135mm with a vintage super takumar 135/3.5 lens. Cheaply acquired on eBay, just some lens fungus to remove and good as new. Right? Here’s the initial result:

Veil nebula with vintage prime lens

This was over 9 hours of RGB data. At the time I was just glad to see the nebula itself in a single frame, but the time I stacked it, all the background stars and chromatic aberration of the old lens rendered the nebula virtually invisible. 

 

Undaunted, I set out to improve upon this image. I picked up an optolong l-enhance filter at roughly 5X the cost of the whole lens. Suddenly the nebula stood out strong against a dim background of stars. But another 9 hours of subs left some great OIII data, but after learning to use the “extract H-a” and OIII debayer algorithms in APP, the OIII data looked great, but the H-a data was complete mush. 
 

I set out with the theory that the lens wasn’t able to focus the two wavelengths at the same point, so I scribed a scale on the lens, and the next clear night I adjusted focus and took test shots, and used the FWHM measurement in APP along with the “extract H-a” option to find out which focus position resulted in the lowest FWHM for the H-a band pass. Success! I could get sharp H-a data at the expense of the OIII band pass. Unfortunately my dew heater failed that night, and I only ended up with about 2 hours of H-a data. 
 

Anyways, this image is my attempt at putting lipstick on a pig. Because of the poor SNR of the H-a data, I couldn’t make a decent HaORGB image. So this is the result of the following:

 

Stack the RGB and the H-a and OIII data separately. Use starnet++ to remove the stars from each channel. 
Since the OIII stars were the smallest, use the stars-only image of the OIII stars as a mask for the RGB stars to remove the CA. 

Start with the star-removed RGB image with the lowest noise in the background. Due to the huge number of large stars, the background was splotchy so apply lots of noise reduction. 
With an inverse luminance mask, blend the sharp HOO nebula data (with stars) into the noise reduced RGB data. 
Add back the focused RGB stars. Lower the opacity to 10% with the “lighten” blending mode in GIMP. Since the sharp HOO image still had stars (starnet++ made mush of it when the stars were removed), this blended the nebula with a reduced star field. 
Add another copy of the RGB stars with the “erode” filter applied, but blended at 100% opacity to bring back the brightest stars. 

Adjust levels/curves/saturation, etc. 

 

This final image isn’t great, but made me learn a lot about post processing. Bracken’s book made for great instruction. Anyways, maybe it’s a bit of an improvement? Suggestions for how I can do this are more than welcome. 
 

All this because of bad optics...

Veil Nebula complex with vintage lens and dual-band filter


  • Starman27, siriusandthepup, psandelle and 11 others like this

#2 John Tucker

John Tucker

    Apollo

  • *****
  • Posts: 1,022
  • Joined: 03 Mar 2018

Posted 05 December 2019 - 11:25 PM

Best post of 2019!  Keep at it!


  • retroman2 and B 26354 like this

#3 the Elf

the Elf

    Vanguard

  • *****
  • Posts: 2,035
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 06 December 2019 - 01:41 AM

Excellent work. I don't think you put lipstik on a pig. I own the 200mm version, the first generation with M42 thread. I use an adaptor to 2'' that is in the lenses filter thread and put a 2'' Baader 8nm O-III an Baader 7nm Ha in front of it. This with a mono modded (bayer pattern removed) T3i plus color data with a stock T3i. The only trouble I found was the red channel of RGB was out of focus due to longitudinal CA but I could partly correct it by deconvoluting the channels seperately. Here is the result:

 

http://www.elf-of-lo...j_Veil2019.html

 

Mind the link to HOO data on the page.

As a comparison to what does not work well here is an attempt with my flat field refractor that has too small a field of view. I put the 0.67 reducer that is designed for RCs behind it and got astigmatism everywhere. A reducer designed for the scope is extremly expensive, more than half the price of the scope and 0.8 only. This is a monochrome Ha image. I colored the nebula in post processing an kept the stars white:

http://www.elf-of-lo...j_Veil2018.html  Check the star shape in the full res. I got a sky full of plus signs!

 

You probably know that a color camera is eating up most of the signal, but if you do Ha and O-III with single filters you'll be fine. It looks like you stopped down the lens, creating spikes at the bright star. If you leave it open and reduce the front lens by an aperture ring (or the only 2'' wide filters) you avoid the spikes.


  • retroman2 and DubbelDerp like this

#4 james7ca

james7ca

    Cosmos

  • *****
  • Posts: 7,850
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 06 December 2019 - 02:52 AM

Some people like star spikes but I'd second the suggestion to use a front mounted step-down filter ring to reduce the aperture of the lens. That will sharpen up the focus and reduce (somewhat) the chromatic aberration and eliminate the star spikes.

 

Depending upon your levels of light pollution you might also want to try imaging with just Ha and OIII without any RGB. Then, of course, you could think about doing a mono modded camera or just switch to a cooled, mono, astro-dedicated camera.

 

Small format astro cameras like the ASI178MM or QHY5III-178M will work very well with short focus, vintage camera lenses. Linked below is an image of the Veil Nebula that I did from a red/orange zone using an old/used Nikkor AI-S 105mm lens with Baader Ha and OIII filters and a ZWO ASI178MM-Cool camera.

 

  https://www.cloudyni...a/#entry8858454

 

All that said, I think you got a pretty nice image given all of your equipment obstacles.


  • retroman2 and DubbelDerp like this

#5 DubbelDerp

DubbelDerp

    Messenger

  • *****
  • topic starter
  • Posts: 422
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 06 December 2019 - 10:37 AM

Excellent work. I don't think you put lipstik on a pig. I own the 200mm version, the first generation with M42 thread. I use an adaptor to 2'' that is in the lenses filter thread and put a 2'' Baader 8nm O-III an Baader 7nm Ha in front of it. This with a mono modded (bayer pattern removed) T3i plus color data with a stock T3i. The only trouble I found was the red channel of RGB was out of focus due to longitudinal CA but I could partly correct it by deconvoluting the channels seperately. Here is the result:

 

http://www.elf-of-lo...j_Veil2019.html

 

Mind the link to HOO data on the page.

As a comparison to what does not work well here is an attempt with my flat field refractor that has too small a field of view. I put the 0.67 reducer that is designed for RCs behind it and got astigmatism everywhere. A reducer designed for the scope is extremly expensive, more than half the price of the scope and 0.8 only. This is a monochrome Ha image. I colored the nebula in post processing an kept the stars white:

http://www.elf-of-lo...j_Veil2018.html  Check the star shape in the full res. I got a sky full of plus signs!

 

You probably know that a color camera is eating up most of the signal, but if you do Ha and O-III with single filters you'll be fine. It looks like you stopped down the lens, creating spikes at the bright star. If you leave it open and reduce the front lens by an aperture ring (or the only 2'' wide filters) you avoid the spikes.

Thanks! That’s a very nice image you got as well. I really like the color and detail you pulled out of the dimmer parts of the nebula. I was fighting noise in the Ha channel due to the short integration, so that limited what I could get from it. 
 

My camera is an Ha-mod t3i, but I was surprised that I still got a good amount of the deep red with only a couple of hours. I also have the 200mm version, and I had initially planned on waiting until next summer to get more Ha at 200mm, but thought I could use this as a learning opportunity to get familiar with masks and layers. 
 

The dual band filter works pretty well with these older lenses as an OIII filter. Focusing with a bahtinov mask was easy since the blue/green is so much brighter. I’d like to get a straight-up Ha filter, though, so I can be a bit more accurate on focusing this wavelength. With the 200mm lens, the 48mm filter would be a nice stop-down, but unfortunately with the 135mm, the front element is only 49mm so doesn’t give much opportunity to stop down without using the aperture blades. I didn’t think about the larger front element on the 200mm... that sounds like a pretty good solution instead of using the internet aperture!



#6 DubbelDerp

DubbelDerp

    Messenger

  • *****
  • topic starter
  • Posts: 422
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 06 December 2019 - 10:44 AM

Some people like star spikes but I'd second the suggestion to use a front mounted step-down filter ring to reduce the aperture of the lens. That will sharpen up the focus and reduce (somewhat) the chromatic aberration and eliminate the star spikes.

 

Depending upon your levels of light pollution you might also want to try imaging with just Ha and OIII without any RGB. Then, of course, you could think about doing a mono modded camera or just switch to a cooled, mono, astro-dedicated camera.

 

Small format astro cameras like the ASI178MM or QHY5III-178M will work very well with short focus, vintage camera lenses. Linked below is an image of the Veil Nebula that I did from a red/orange zone using an old/used Nikkor AI-S 105mm lens with Baader Ha and OIII filters and a ZWO ASI178MM-Cool camera.

 

  https://www.cloudyni...a/#entry8858454

 

All that said, I think you got a pretty nice image given all of your equipment obstacles.

Thanks for the suggestions! And that’s a really nice image you linked. As mentioned above, the 135mm lens doesn’t offer much opportunity to stop the lens down with a 48mm filter, but I really like the idea of trying again with the 200mm lens with the aperture blades wide open. I’m not particularly fond of diffraction spikes, but I left that one in because starnet++ really obliterated the starless image in those areas. Not the fault of the program, but mediocre data with diffraction spikes. I thought the HOO with stars image kept the best detail, even though the stars were so dominant. 

 

My skies are around Bortle 3.5-ish. Pretty dark, so I thought I could get away with RGB at first. I didn’t care for the colors of the stars with the HOO version, and the background was so noisy that I went back to the other image for the background and star field. Although I don’t have much star color regardless... I think the more neutral colored RGB stars made for a better overall image. 



#7 james7ca

james7ca

    Cosmos

  • *****
  • Posts: 7,850
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 06 December 2019 - 04:41 PM

You don't need to stop or start with a 48mm step down or filter. You can stack step-down rings to get a broad range of effective f-ratios. You may also be able to find some step-down rings with pretty large steps, like this one (on Amazon) that goes from 49mm down to 37mm. I'm not sure what filter size is used on your 135mm f/3.5 (maybe 49mm?), but you could mount something like the 49-37mm on your lens and then mount a 37-30mm step-down in that to produce an aperture of 30mm (f/4.5). Here is a link to an inexpensive step-down ring set that includes pieces that go down to 26mm (≈f/5.2 on your 135mm lens):

 

  https://www.amazon.c...U_ByS6DbX5B1EG3

 

You can nest these rings and in some cases mount them backwards (in the previous ring) so that the limiting aperture stays close to the front lens element (you want to do that to prevent vignetting). So, if your lens used 49mm filters you could start with the two rings that give you 30mm and then mount the 30-26mm step down backwards in the 37-30mm ring. That would put the 26mm aperture almost at the normal filter location, just in front of the lens.

 

You'll also want to get a set of step-up rings, since you can use those in combinations with the step-downs.

 

That said, these kinds of combinations tend to work better on slightly longer focal length lenses, although I've used this technique on lenses as short as 50mm. Also, B&H Photo seems to have a slightly larger range of rings than you will find on Amazon.


Edited by james7ca, 06 December 2019 - 04:42 PM.


#8 the Elf

the Elf

    Vanguard

  • *****
  • Posts: 2,035
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 07 December 2019 - 03:07 AM

The 200mm Takumar did not quite come to focus with the Ha filter on. I found this refurb video:

https://www.youtube....h?v=uA5GW-ba3Jc

(Make sure speakers are off! Awful music, the guy does not speak a word.)

Just after first few steps you get access to the infinity stop meachanism of the focus ring. I adjusted it so that I can turn the focus ring a bit further than infinity now. It is quite self explaining when you see it. Now I have good focus with the Ha filter as well. It was not more than about 10 min work.


Edited by the Elf, 07 December 2019 - 03:09 AM.


#9 gene williams

gene williams

    Vostok 1

  • *****
  • Posts: 174
  • Joined: 17 Jan 2011
  • Loc: Western Missouri

Posted 07 December 2019 - 10:14 AM

 

Well this was an image that just about drove me mad. As a complete AP beginner, with a freshly acquired camera tracker, I set out to capture the veil nebula complex at 135mm with a vintage super takumar 135/3.5 lens. Cheaply acquired on eBay, just some lens fungus to remove and good as new. Right? Here’s the initial result:

 

This was over 9 hours of RGB data. At the time I was just glad to see the nebula itself in a single frame, but the time I stacked it, all the background stars and chromatic aberration of the old lens rendered the nebula virtually invisible. 

 

Undaunted, I set out to improve upon this image. I picked up an optolong l-enhance filter at roughly 5X the cost of the whole lens. Suddenly the nebula stood out strong against a dim background of stars. But another 9 hours of subs left some great OIII data, but after learning to use the “extract H-a” and OIII debayer algorithms in APP, the OIII data looked great, but the H-a data was complete mush. 
 

I set out with the theory that the lens wasn’t able to focus the two wavelengths at the same point, so I scribed a scale on the lens, and the next clear night I adjusted focus and took test shots, and used the FWHM measurement in APP along with the “extract H-a” option to find out which focus position resulted in the lowest FWHM for the H-a band pass. Success! I could get sharp H-a data at the expense of the OIII band pass. Unfortunately my dew heater failed that night, and I only ended up with about 2 hours of H-a data. 
 

Anyways, this image is my attempt at putting lipstick on a pig. Because of the poor SNR of the H-a data, I couldn’t make a decent HaORGB image. So this is the result of the following:

 

Stack the RGB and the H-a and OIII data separately. Use starnet++ to remove the stars from each channel. 
Since the OIII stars were the smallest, use the stars-only image of the OIII stars as a mask for the RGB stars to remove the CA. 

Start with the star-removed RGB image with the lowest noise in the background. Due to the huge number of large stars, the background was splotchy so apply lots of noise reduction. 
With an inverse luminance mask, blend the sharp HOO nebula data (with stars) into the noise reduced RGB data. 
Add back the focused RGB stars. Lower the opacity to 10% with the “lighten” blending mode in GIMP. Since the sharp HOO image still had stars (starnet++ made mush of it when the stars were removed), this blended the nebula with a reduced star field. 
Add another copy of the RGB stars with the “erode” filter applied, but blended at 100% opacity to bring back the brightest stars. 

Adjust levels/curves/saturation, etc. 

 

This final image isn’t great, but made me learn a lot about post processing. Bracken’s book made for great instruction. Anyways, maybe it’s a bit of an improvement? Suggestions for how I can do this are more than welcome. 
 

All this because of bad optics...

 

Very nice, and I see no "lipstick on a pig" lol.  If I may ask, how did you mount the dual band filter?  I am interested in trying this technique.



#10 DubbelDerp

DubbelDerp

    Messenger

  • *****
  • topic starter
  • Posts: 422
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 07 December 2019 - 06:40 PM

Very nice, and I see no "lipstick on a pig" lol.  If I may ask, how did you mount the dual band filter?  I am interested in trying this technique.

Thanks! It’s pretty simple to attach. The optolong filter has an m48x.75 thread, which is compatible with lens filter threads. So it was just a step down ring from the diameter of the lens filter thread to 48mm. In this case, the lens has a 49mm filter thread, so a 49-48mm step-down ring was all I needed. 


  • gene williams likes this

#11 DubbelDerp

DubbelDerp

    Messenger

  • *****
  • topic starter
  • Posts: 422
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 07 December 2019 - 08:35 PM

Astrobin link, if anyone is interested...
https://www.astrobin...u12nl1/?nc=user

#12 DubbelDerp

DubbelDerp

    Messenger

  • *****
  • topic starter
  • Posts: 422
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 07 December 2019 - 08:40 PM

The 200mm Takumar did not quite come to focus with the Ha filter on. I found this refurb video:

https://www.youtube....h?v=uA5GW-ba3Jc

(Make sure speakers are off! Awful music, the guy does not speak a word.)

Just after first few steps you get access to the infinity stop meachanism of the focus ring. I adjusted it so that I can turn the focus ring a bit further than infinity now. It is quite self explaining when you see it. Now I have good focus with the Ha filter as well. It was not more than about 10 min work.

I had the same problem with mine. But I had to fully disassemble mine since the rear element had fungus. Really seems like this lens is susceptible to the rear lens getting contaminated... I’ve owned three, and two have had lens fungus on the rear element only. I messed up the reassembly, and stumbled across the readjustment of the infinity point. But it works great, and since I bought it with lens fungus, it cost less than $25 USD! 
 

I have to say, it’s REALLY satisfying to bring an old lens back to life, and save it from the landfill. Even with the CA, it’ll be a perfect narrowband lens moving forward in its new life. 



#13 the Elf

the Elf

    Vanguard

  • *****
  • Posts: 2,035
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 08 December 2019 - 07:32 AM

https://www.teleskop...aobjektive.html

 

This ring fits the 200mm Takumar and takes standard 2'' filters. When the filters have a front thread you can add a 2'' extension like 40mm or so as a dew shield in front. I always do:

 

Tak_low.JPG

 

Stepped down to 48mm I always leave the internal aperture full open and thus have no spikes at stars:

http://www.elf-of-lo...k_2019_full.jpg

http://www.elf-of-lo...l_2019_1080.jpg

http://www.elf-of-lo...n_2019_full.jpg


Edited by the Elf, 08 December 2019 - 07:37 AM.



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