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

How to improve?

  • Please log in to reply
15 replies to this topic

#1 DrJones

DrJones

    Lift Off

  • -----
  • topic starter
  • Posts: 19
  • Joined: 02 Feb 2019

Posted 26 June 2019 - 04:11 PM

Hi together,

 

I finally got my setup complete and started imaging. Since I'm doing this from the suburbs of the SF peninsula, light pollution is present but has to work for now.

 

My Setup:

- Mount Skywatcher EQ6R-Pro

- ExploreScientific ED127-100

- Now ASI1600mm-Pro with the 8x Zwo filter set

- Guide scope Astromania 60mm, guide cam Zwo Asi120mm-s and a dedicated RapsperrbyPi3+ for operating PHD2

 

I acquired 10x2min per channel for Ha, O3 and SII with gain 138 using SequenceGenerator at -20C camera temp, registered with Fitswork (and also tested registration with DeepSkyStacker) and did the false-color overlay with Photoshop.

 

I have a feeling that the image is still somewhat grainy - is that due to the gain setting I am using (unity is supposed to be at 139) or do I have collect more images (integrate for a longer time)? Is 2 min fine or would I run into issues with light pollution? Do you have any suggestions?

 

Sorry, I'm new, still learning... is there a better way of doing (the correctly registered) overlays? 

 

Thanks for any input!

Attached Thumbnails

  • M17.jpg

  • elmiko, epdreher, 2ghouls and 2 others like this

#2 kyle528

kyle528

    Messenger

  • *****
  • Posts: 473
  • Joined: 02 Feb 2019
  • Loc: SE Ohio

Posted 26 June 2019 - 04:19 PM

Overall integration time is king. Look at what only 1 hour gave you, imagine what 5, or 10 or 15 hours would look like. Something I had to learn early on in my AP experience was that what we are doing here should really be treated not as snapshots, but as projects. YMMV, but I discovered quite quickly the value of many hours of integration, and the effect is has on the end result. I shoot OSC, so I can not offer any advice on narrowband techniques. Overall, a great result for only 1 hour! Aligning the channels may fix that purple background


  • bobzeq25, ks__observer, OldManSky and 1 other like this

#3 ButterFly

ButterFly

    Mariner 2

  • *****
  • Posts: 287
  • Joined: 07 Sep 2018

Posted 26 June 2019 - 04:54 PM

This is excellent work for 20 minutes of integration time per channel.  You increase signal to noise ratio by sqrt(N) with the number of stacks.  Change one thing at a time to see how your setup works under your conditions.  So ... add more subs to reduce graininess first.  Doubling the integration time will yield a 45% increase in SNR, assuming similar seeing conditions.   Your scope with your camera ( I am assuming no focal reducer) is great for well sampled images in 1-2" seeing.

 

To get an idea of the considerations involved with optimal sub-exposure length based on light pollution, focal length ..., watch this video: Robin Glover's recent youtube video presentation as well as the addedum on gain settings.

 

Total integration time is very important and the breakdown of number of subs vs sub length can help optimize your SNR.

 

Gain depends heavily on your target.  You want to utilize as many bins as possible for your target.  When there are other things in the field that can swamp, you can always cut it out and replace that area with less swamped images for that area.


  • DrJones likes this

#4 nimitz69

nimitz69

    Apollo

  • *****
  • Posts: 1469
  • Joined: 21 Apr 2017
  • Loc: A barrier island 18 miles south of Cocoa Beach

Posted 26 June 2019 - 04:55 PM

Are you taking calibration frames? (Darks, flats & bias or dark flats).  If not you should start doing this immediately and for every iamaging session, it will make a big difference in the final image Assuming you’ve otherwise collected good data ... developing good imaging habits early on is key to producing quality images and if you skip things now you’re likely to start cutting corners elsewhere as well. ...


  • ks__observer and DrJones like this

#5 DrJones

DrJones

    Lift Off

  • -----
  • topic starter
  • Posts: 19
  • Joined: 02 Feb 2019

Posted 27 June 2019 - 07:46 PM

Hi all, thanks for the replies!

 

I am taking flat frames after each session and I took dark frames.

 

I guess the trick here really is getting more light as it seems. I will try to see what happens when you add another h or two.

 

I typically have a window of 2 hours for each object... any advice on how to best register images from different sessions/days?

 

Is everyone using Pixelinsight for the the aligning or can you recommend other software? I don't have to have a nice interface, I can work my way through anything provided there is some guideline. Are the astro plugins for Fiji?

 

Thanks for the video, I will educate myself a bit more with this one! I shot the same object with LRGB filters and have to say that the difference with using narrowband filters is tremendous!



#6 Ballyhoo

Ballyhoo

    Soyuz

  • *****
  • Posts: 3941
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 27 June 2019 - 08:14 PM

Just curios if you are coming here to read kudos? are you playing CN for praise? Of course  it's beautiful and you must know that.  OR are you like Pablo Picasso wanting to cut his ear?



#7 fewayne

fewayne

    Viking 1

  • *****
  • Posts: 539
  • Joined: 10 Sep 2017
  • Loc: Madison, WI, USA

Posted 27 June 2019 - 08:41 PM

Isn't narrowband amazing for us bright-sky folks? Suddenly I are almost a real astronomer, I are!

 

If you can figure out a somewhat abstruse interface, open-source Siril does a good job at the basics of registration and stacking. You can do RGB composition there, as well as stretching (there's a one-button autostretch tool). It has tools for background extraction, color calibration, and much more, but I never got good enough with it to consistently get results I liked. You will definitely want to hunt up some tutorials because at least for me, it wasn't at all obvious how to proceed.

 

I used to have trouble with getting good color out of my DSLR images with Siril, but that's less of an issue now with narrowband.

 

I recently tried Astro Pixel Processor for a month, and wound up buying it. It does not have as many features (yet) as PixInsight, but from what I can see it is a great deal easier to use, and does a really excellent job at processing images with a comprehensible workflow. 130 euros instead of 250.


  • DrJones likes this

#8 DrJones

DrJones

    Lift Off

  • -----
  • topic starter
  • Posts: 19
  • Joined: 02 Feb 2019

Posted 27 June 2019 - 08:50 PM

No, I am actually not looking for praise of my work also it's nice that people like this. I'm looking for honest feedback...and compared to the masterpieces I've seen posted in this forum I think there is quite a lot of room for improvement!

 

Specifically: 

- In the attached image you can clearly see that the stars are not perfectly aligned (and I have no idea on how to do this properly).

- You can clearly see the noise (this has been addressed previously by people mentioning to gather more light (double the time you get a 45% better signal to noise-rule by ButterFly).

- I think the way I combine the colors is not correct (Ha - red, O3 - green, SII blue). Do I need to add R, G, B or Lum? Having a look at the Hubble palette... how do I do this with photoshop and not PixInsight?

- Working with the ASI1600mm-pro - is a gain of 138 and 2 min exposure really the way to go?

- Do I have to re-focus for every filter or should I assume that all wavelengths have the same focus?

 

Thanks for the feedback so far!

Attached Thumbnails

  • M17_crop.jpg


#9 View2

View2

    Vanguard

  • *****
  • Posts: 2037
  • Joined: 20 May 2016
  • Loc: Vancouver, WA USA

Posted 27 June 2019 - 08:54 PM

More subs

#10 Monkeybird747

Monkeybird747

    Viking 1

  • *****
  • Posts: 562
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 27 June 2019 - 08:55 PM

Hi all, thanks for the replies!

 

I am taking flat frames after each session and I took dark frames.

 

I guess the trick here really is getting more light as it seems. I will try to see what happens when you add another h or two.

 

I typically have a window of 2 hours for each object... any advice on how to best register images from different sessions/days?

 

Is everyone using Pixelinsight for the the aligning or can you recommend other software? I don't have to have a nice interface, I can work my way through anything provided there is some guideline. Are the astro plugins for Fiji?

 

Thanks for the video, I will educate myself a bit more with this one! I shot the same object with LRGB filters and have to say that the difference with using narrowband filters is tremendous!

 

I've only used Pixinsight, so I can't give any comparisons to other processing software. I can only say give it a look if you haven't already, as it is very powerful and has a large user base for support. When I started imaging I started with PI, and I have no regrets.

 

As for how to improve, can you post up a link to full res version? You might get more useful feedback that way, as it's difficult to judge from the jpg. Narrowband color mapping is in the eye of the beholder. If you like your overall colors, then I'd start working on masking the stars and getting the purple and green out of them. Learning to generate effective, specialized masks made a huge impact on my processing. That learning is still in progress. 


  • DrJones likes this

#11 kyle528

kyle528

    Messenger

  • *****
  • Posts: 473
  • Joined: 02 Feb 2019
  • Loc: SE Ohio

Posted 27 June 2019 - 09:05 PM

....huh?

ohmy.gif

Also, van Gogh...


I don’t get it either... Not sure what the OP did to have that odd string of words thrown at them


Sent from my iPhone using Tapatalk

#12 Monkeybird747

Monkeybird747

    Viking 1

  • *****
  • Posts: 562
  • Joined: 31 Aug 2017
  • Loc: Louisville, KY

Posted 27 June 2019 - 09:09 PM

I don’t get it either... Not sure what the OP did to have that odd string of words thrown at them


Sent from my iPhone using Tapatalk

Deleted. Must not engage...;)


  • kyle528 likes this

#13 ButterFly

ButterFly

    Mariner 2

  • *****
  • Posts: 287
  • Joined: 07 Sep 2018

Posted 28 June 2019 - 04:51 AM

- I think the way I combine the colors is not correct (Ha - red, O3 - green, SII blue). Do I need to add R, G, B or Lum? Having a look at the Hubble palette... how do I do this with photoshop and not PixInsight?

- Working with the ASI1600mm-pro - is a gain of 138 and 2 min exposure really the way to go?

 

 

These two questions are related.  The answer is that no one can find out other than you just messing around.  It is better to guide your trials rather than just shoot in the dark.  Those lectures went over some of the considerations, but the biggest variable unknown to us is the light pollution in your area, and particularly, its distribution.  Get together with locals for the best guides in that arena, or just try to find out yourself.

 

Lum is essentially no filter at all.  It catches everything like starlight, but also ALL your light pollution at ALL of its wavelengths.  RGB are rather broad, but have cutoffs.  They let only some band of your light pollution through, but it's a fairly large band.  The narrowband filters are much more narrow, letting only small wavelength ranges through.  Even though you now have only a tiny band of light pollution getting through, you also have only a tiny band of signal getting through.  If the light pollution level in that band is pretty much the same as the signal in that band, your job is very difficult and you need very long total integration times.  The broader the band, the more total light pollution hitting the pixels, so shorter subs lengths compared to narrow bands.  At lum, 2 mins may be way too much.  Ideally, you choose the bands based on the target.  If a particular nebula is emitting in OIII and not very much Ha, then it's just a waste of time to image that Ha.  If it's a reflection nebula that is refecting the blackbody curve of the illuminating star, perhaps narrowbands are better than RGB because of your local light pollution.  If you want to mix the narrowbands with the RGB, you will need to factor how the local light pollution is in those bands to adjust the sublengths and total integraton time for each band.  "Correct" at the end of the day only means that you enjoyed doing what you did.

 


- Do I have to re-focus for every filter or should I assume that all wavelengths have the same focus?

 

This one is fairly easy to check.  It depends on the scope and how you put the filter in the train (wheel vs. screwed onto camera).  Just focus on a star in the center in one band using a Bahtinov mask then switch bands.  Any focus shift will be apparent.  Arcturus and Vega are both fairly high now and have very different spectral profiles in the visible band.  Any shift you notice should remain the same night to night.


  • DrJones likes this

#14 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 991
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 28 June 2019 - 06:02 AM

Re star alignment in PI -- i believe you have to use Dynamic Alignment after you stack align to fine tune.



#15 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 16931
  • Joined: 27 Oct 2014

Posted 28 June 2019 - 09:15 AM

Hi all, thanks for the replies!

 

I am taking flat frames after each session and I took dark frames.

 

I guess the trick here really is getting more light as it seems. I will try to see what happens when you add another h or two.

 

I typically have a window of 2 hours for each object... any advice on how to best register images from different sessions/days?

 

Is everyone using Pixelinsight for the the aligning or can you recommend other software? I don't have to have a nice interface, I can work my way through anything provided there is some guideline. Are the astro plugins for Fiji?

 

Thanks for the video, I will educate myself a bit more with this one! I shot the same object with LRGB filters and have to say that the difference with using narrowband filters is tremendous!

You need to take bias (or, with a 1600, dark flats), also.  Otherwise the mathematics for flat correction don't work (because it's multiplicative, and the signal levels are so different).

 

Are you using the same reference frame for all filters?

 

PixInsight registers as well as anything.  Deep Sky Stacker is "good enough".  My recommendation for a beginner (and it's good enough for some experienced imagers), is Astro Pixel Processor.   PI has a small edge, but only if you learn to adjust the many parameters well.  I have hundreds of hours in that.

 

Here's the "guideline" for PI.  It's fairly basic.  Good enough to do some fine images, but there's only so much you can do with PI in 450 pages.  <smiling, but not kidding>

 

https://www.amazon.c.../dp/3319976885/

 

People who are not experienced in PI sometimes complain about the $250 cost.  People who are know it's insignificant compared to the cost in your time.

 

You image over multiple nights by using platesolving to get to almost exactly the same place, then alignment in processing makes things exact to the fraction of a pixel.  Platesolving is computerized pattern recognition of a test image that tells you exactly (understatement) where you're pointed.  I like PlateSolve2, this is one of the easier things to learn.


Edited by bobzeq25, 28 June 2019 - 09:27 AM.

  • kyle528 and DrJones like this

#16 2ghouls

2ghouls

    Viking 1

  • *****
  • Posts: 794
  • Joined: 19 Sep 2016

Posted 28 June 2019 - 09:31 AM

No, I am actually not looking for praise of my work also it's nice that people like this. I'm looking for honest feedback...and compared to the masterpieces I've seen posted in this forum I think there is quite a lot of room for improvement!

 

Specifically: 

- In the attached image you can clearly see that the stars are not perfectly aligned (and I have no idea on how to do this properly).

- You can clearly see the noise (this has been addressed previously by people mentioning to gather more light (double the time you get a 45% better signal to noise-rule by ButterFly).

- I think the way I combine the colors is not correct (Ha - red, O3 - green, SII blue). Do I need to add R, G, B or Lum? Having a look at the Hubble palette... how do I do this with photoshop and not PixInsight?

- Working with the ASI1600mm-pro - is a gain of 138 and 2 min exposure really the way to go?

- Do I have to re-focus for every filter or should I assume that all wavelengths have the same focus?

 

Thanks for the feedback so far!

-It's not clear from the image sample if the problem is truly alignment or other processing steps. For example, different star spot sizes due to different stretch on each narrowband master. I would recommend registering all your narrowband frames to one reference frame if you aren't already doing that.

 

-Noise. Yeah, I agree. The answer is almost always more integration.

 

-There is no correct way, but you did HOS, hubble palette is SHO. Here is a classic tutorial on SHO processing in Photoshop by Bob Franke: http://bf-astro.com/hubbleP.htm

 

-I do gain 200 and 5 min exposures for narrowband, but it depends on your filter bandpass and sky conditions. I think if you are hitting somewhere between 1500 and 5000 median ADU and not clipping too many star cores, you should be fine.

 

-Yes, most likely you should be re-focusing for every filter change, and even more importantly throughout the night due to temperature shifts. I have parfocal Astrodon filters, and a fairly high-end, well-corrected refractor and still have to re-focus when I ever move from red to blue or Ha to OIII because there is a focus shift. Very little between Lum, Red, and Green however. Easiest way to test if you don't have automated focusing is with a bahtinov mask.


  • DrJones 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