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

A dead easy Andromeda Galaxy (M31)

  • Please log in to reply
32 replies to this topic

#1 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 13 October 2021 - 09:39 AM

Hello all,

 

every now and then I read warnings that Andromeda is a difficult target. You should take images with different exposure times to save the faint details, the core is very bright and likely saturated and so on. I'd like to share a very simple approach. I want only one sub exposure time and simply take plenty of them. The sub length supposed to be as long as possible but short enough to avoid saturation. Surprising fact: the smaller M32 comes with a brighter core than the large M31. So this is the limiting factor. There are a few bright stars around. Depending on the FOV you probably have the mag 4.5 nu Andromedae in the frame (upper right corner in my image). It is a deep blue B5V star. It is visible to the naked eye so we have to accept saturation here. Using my 65mm quad at f/6.5 and an unmodded Canon EOS 800D (T7i) a 3 minute sub exposure time leads to these level readings:

M31 core: 0.43
M32 core: 0.65

----

EDIT: ISO 400 was used.

----

In both cases the red channel happened to be the brightest. Saturation is near 1.0 (it is not exactly 1.0, nu And reads 0.976). Are there any other saturated stars? Yes, there are 35 saturated stars out of 31015 stars that PixInsight detects in my stack. That is about 1 in a thousand. Seems acceptable to me. Here is a map of the saturated stars:

 

35_sat_stars.png

 

Now that the nights are getting longer I was able to capture 160 subs = 8 hours of data over a 9 hour period. The missing hour is due to meridian flip and checking focus twice and a few seconds for dither. I dithered every sub. As expected the early evening and the early morning shots when the object was lower in the sky are a bit blurred. After sorting out all subs with an FWHM larger than 4 arcseconds (that is 2.2 pixels at 1.814 arcsec/pix) there were 120 subs equals exactly 6 hours left for the stack. And this is all the magic. No filters, no different exposure times, no luminance, no modded camera, all dead simple. And this is the result:

 

M31_2021_low.jpg

 

link to full res (15MB): https://elf-of-lothl...1_2021_full.jpg

 

My sky background is approx. Bortle 4. My conclusion is that Andromeda is not as demanding as sometimes told as far as image acquisition is concerned. Sure, my image does not show all the faint details. My original plan was to capture two nights and go with 10h at least but the clouds did not offer this opportunity. As the full moon is close I decided to process the 6h and I am not disappointed.

 

C&C welcome. Anyone is invited to post his or her image and point out what can be done better without making matters too complicated. Maybe you can make this thread kind of a Andromeda-how-to. Take my considerations as a starting point for a _friendly_ discussion, please!

 

CS

the Elf


Edited by the Elf, 13 October 2021 - 01:21 PM.

  • nofxrx, DaveD, Phillip Creed and 27 others like this

#2 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 13 October 2021 - 09:53 AM

Here is a link to the stack if anyone wants to play with it. File size is 291MB.

http://elf-of-lothlo...20x3min_OSC.tif


  • galacticinsomnia likes this

#3 matt_astro_tx

matt_astro_tx

    Surveyor 1

  • -----
  • Posts: 1,743
  • Joined: 19 Jan 2021
  • Loc: Dallas, Texas

Posted 13 October 2021 - 10:56 AM

Fantastic image.  I'm shooting M31 now as well.  Up to about 8 hours I think.  I'm shooting for ~15-20 hours to really pull out the color.  My processing skills just aren't up to task with the amount of data I have now, I guess.


Edited by matt_astro_tx, 13 October 2021 - 12:19 PM.


#4 bobzeq25

bobzeq25

    ISS

  • *****
  • Posts: 25,909
  • Joined: 27 Oct 2014

Posted 13 October 2021 - 10:56 AM

That's quite good.  What can make this a hard target for a beginner is the large dynamic range.


  • Ivo Jager and Benschop like this

#5 Steve OK

Steve OK

    Surveyor 1

  • -----
  • Posts: 1,857
  • Joined: 22 Sep 2007
  • Loc: OKC, OK

Posted 13 October 2021 - 10:59 AM

Beautiful color there!  Dead simple, maybe, but 8 hours is nothing to sneeze at.  I think we can get swamped by the idea that this has to be intricately complicated to get good results.  This image shows that is not always true.  Nice work.

 

Steve



#6 Alen K

Alen K

    Surveyor 1

  • *****
  • Posts: 1,728
  • Joined: 25 Nov 2009

Posted 13 October 2021 - 11:28 AM

That's a really nice image. I like the colors. Did you do any star reduction or any steps that limited their size? IMO they look nicely subdued without being too small. What ISO did you use? 

 

It would seem everyone is working on their M31 images right now and I am no exception. Although I have tried it with telephoto lenses up to 300mm and in the ancient past on 35mm film with a 762mm focal-length telescope (fit the frame nicely), this is the first time I have seriously tried it with a DSLR. But I am using an APS-C sensor and it no longer fits the frame nicely at 762mm focal length. (Yup, same telescope I used with 35mm.) So I have resorted to doing a four-panel mosaic. A lot more work but I am getting a similar field of view after cropping as what you are showing. I used 30-second, unguided exposures at ISO 1600 on a Pentax K-3II and likewise see only a few stars that mildly saturated. (I use the camera's highlight alert on previews to check this at the telescope.) One difference is that I shot with M31 oriented in the frame the other way diagonally and hence didn't get nu Andromedae in that relevant panel, which otherwise would have indeed saturated.    

 

However, I am only up to a little over 2.5 hours of data. I hope to add a couple of hours this new-moon period if the weather permits. I don't think I will get six hours on M31 this year since I would prefer to use the opportunity (assuming I do get it) to do winter stuff like M42 and the Horsehead in the morning hours. I don't normally have access to a dark sky site in the dead of winter. 


Edited by Alen K, 13 October 2021 - 11:55 AM.


#7 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 13 October 2021 - 01:21 PM

Thank you all for your kind words!

 

I applied a standard deconvolution using an actual star pattern from DynamicPFS. The FWHM was 2.2 pixels before and 1.9 pixels after deconvolution and denoise. For a DSLR with a spatial low pass 1.5 to 2 pixels is the minimum size you can get. I think single pixel stars are only possible in a mono camera. I an OSC there is no way to determine the color unless drizzle integration is used. I always end up near 2 when imaging with a short focal length. In general, when the scale equals the seeing the star size is what the low pass filters allows it to be. For the lower resolution image uploaded here I uses a tiny bit of unsharp mask to make the stars pop a bit more after the resample.

 

I used ISO 400. Going to add this in the original post. Thanks for asking.


Edited by the Elf, 13 October 2021 - 01:31 PM.

  • bobzeq25 likes this

#8 F.Meiresonne

F.Meiresonne

    Cosmos

  • *****
  • Posts: 7,802
  • Joined: 22 Dec 2003
  • Loc: Eeklo,Belgium

Posted 13 October 2021 - 01:29 PM

Ok, so my go in Startools

 

https://www.cloudyni...11_17091599.png

 

Very well done Elf, as usual.

 

Personaly i find the center a bit, dunno , slight halo sight in it.

I am probably wrong, i really have an issue in evaluating pictures , this not easy and in many cases i seem not to see things right...ah well, it stays fun...

 

Oh yeah, and the data was excellent.

Your capture skills definately are much better then mine, but i also suspect there is less LP, using wipe was hardly needed...

Attached Thumbnails

  • M31 (Large).jpg

Edited by F.Meiresonne, 13 October 2021 - 01:31 PM.

  • Steve OK and galacticinsomnia like this

#9 galacticinsomnia

galacticinsomnia

    Surveyor 1

  • *****
  • Posts: 1,647
  • Joined: 14 Aug 2020
  • Loc: Pacific Northwest - Oregon

Posted 13 October 2021 - 01:36 PM

Hello all,

 

every now and then I read warnings that Andromeda is a difficult target. You should take images with different exposure times to save the faint details, the core is very bright and likely saturated and so on. I'd like to share a very simple approach. I want only one sub exposure time and simply take plenty of them. The sub length supposed to be as long as possible but short enough to avoid saturation. Surprising fact: the smaller M32 comes with a brighter core than the large M31. So this is the limiting factor. There are a few bright stars around. Depending on the FOV you probably have the mag 4.5 nu Andromedae in the frame (upper right corner in my image). It is a deep blue B5V star. It is visible to the naked eye so we have to accept saturation here. Using my 65mm quad at f/6.5 and an unmodded Canon EOS 800D (T7i) a 3 minute sub exposure time leads to these level readings:

M31 core: 0.43
M32 core: 0.65

----

EDIT: ISO 400 was used.

----

In both cases the red channel happened to be the brightest. Saturation is near 1.0 (it is not exactly 1.0, nu And reads 0.976). Are there any other saturated stars? Yes, there are 35 saturated stars out of 31015 stars that PixInsight detects in my stack. That is about 1 in a thousand. Seems acceptable to me. Here is a map of the saturated stars:

 

attachicon.gif35_sat_stars.png

 

Now that the nights are getting longer I was able to capture 160 subs = 8 hours of data over a 9 hour period. The missing hour is due to meridian flip and checking focus twice and a few seconds for dither. I dithered every sub. As expected the early evening and the early morning shots when the object was lower in the sky are a bit blurred. After sorting out all subs with an FWHM larger than 4 arcseconds (that is 2.2 pixels at 1.814 arcsec/pix) there were 120 subs equals exactly 6 hours left for the stack. And this is all the magic. No filters, no different exposure times, no luminance, no modded camera, all dead simple. And this is the result:

 

attachicon.gifM31_2021_low.jpg

 

link to full res (15MB): https://elf-of-lothl...1_2021_full.jpg

 

My sky background is approx. Bortle 4. My conclusion is that Andromeda is not as demanding as sometimes told as far as image acquisition is concerned. Sure, my image does not show all the faint details. My original plan was to capture two nights and go with 10h at least but the clouds did not offer this opportunity. As the full moon is close I decided to process the 6h and I am not disappointed.

 

C&C welcome. Anyone is invited to post his or her image and point out what can be done better without making matters too complicated. Maybe you can make this thread kind of a Andromeda-how-to. Take my considerations as a starting point for a _friendly_ discussion, please!

 

CS

the Elf

What did you use for post. ?

Nice work..

 

Clear Skies !!



#10 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 13 October 2021 - 01:38 PM

Thank you for taking the effort, Freddy. I tried a similar version. When using the auto stretch parameters in HistogramTransformation and applying HDRMultiscaleTransform PI creates about the same look with the dust lanes visible near the core. The downside was a somewhat unnatural background with faint stars so I went back to my usual processing with arcsinh. Your image combines the HDR look with more dynamic but less colored stars. A nice one for sure!

You are right, my core shows a halo from histogram equalization. I'm still seeking a better way. I'm also troubled with the core of globular clusters. Still some room for improvement.


Edited by the Elf, 13 October 2021 - 01:41 PM.

  • F.Meiresonne likes this

#11 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 13 October 2021 - 01:39 PM

What did you use for post. ?
 

All in PixInsight. Pre and Post.


  • galacticinsomnia likes this

#12 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 13 October 2021 - 01:51 PM

Does anyone know how to decrease this terrible background pattern? It seems like the Canons cannot produce an even background.

 

100% crop of upper right corner, stars removes and blurred.

 

BG_pattern.jpg



#13 F.Meiresonne

F.Meiresonne

    Cosmos

  • *****
  • Posts: 7,802
  • Joined: 22 Dec 2003
  • Loc: Eeklo,Belgium

Posted 13 October 2021 - 02:26 PM

Thank you for taking the effort, Freddy. I tried a similar version. When using the auto stretch parameters in HistogramTransformation and applying HDRMultiscaleTransform PI creates about the same look with the dust lanes visible near the core. The downside was a somewhat unnatural background with faint stars so I went back to my usual processing with arcsinh. Your image combines the HDR look with more dynamic but less colored stars. A nice one for sure!

You are right, my core shows a halo from histogram equalization. I'm still seeking a better way. I'm also troubled with the core of globular clusters. Still some room for improvement.

Actually is tempered a bit the color in the stars, by default the colors were much stronger, but i don't always like it, just me...still a sort of looking through the eyepiece syndrome, not much color in stars there, allthough not completely absent either....



#14 vidrazor

vidrazor

    Apollo

  • *****
  • Posts: 1,486
  • Joined: 31 Oct 2017
  • Loc: North Bergen, NJ

Posted 13 October 2021 - 03:04 PM

Does anyone know how to decrease this terrible background pattern? It seems like the Canons cannot produce an even background.

100% crop of upper right corner, stars removes and blurred.

I would process for luma and chroma NR in a separate layer and drop the skies in through a luminosity mask, and carefully adjust the luma transition point. Here's a quickie 100% sample from your full res file (click to view 100%).

Attached Thumbnails

  • luma masking.jpg

Edited by vidrazor, 13 October 2021 - 03:08 PM.

  • the Elf likes this

#15 17.5Dob

17.5Dob

    Fly Me to the Moon

  • *****
  • Posts: 7,388
  • Joined: 21 Mar 2013
  • Loc: Colorado,USA

Posted 13 October 2021 - 04:51 PM

Nice write up ! I have always found Andromeda to be an easy target as well. The dynamic range can be easily captured in a single exposure. As long as you shoot the subs as long as you can without saturating the core during data collection, you'll have more than enough data to pull out the fainter portions as well....

Now, M42 is an entirely different matter.....



#16 galacticinsomnia

galacticinsomnia

    Surveyor 1

  • *****
  • Posts: 1,647
  • Joined: 14 Aug 2020
  • Loc: Pacific Northwest - Oregon

Posted 13 October 2021 - 05:58 PM

This has to be one of the best data sets I have seen for this target.  Really nice work !

As far as a flatter field.  I would take a strong look at Siril Background removal, or the amazing Affinity Photo for neutralzing some of the noise with its background removal/calibration tools. 
It has some amazing features in setting levels, masking, and gamma. 

Looks great regardless.

Clear Skies !!



#17 Tim J Fowler

Tim J Fowler

    Viking 1

  • *****
  • Posts: 579
  • Joined: 21 Jul 2020
  • Loc: Seneca, South Carolina

Posted 13 October 2021 - 07:50 PM

Another great image as usual, Elf. I tried your link to the stack but for some reason I get nothing. It'll open another window and then automatically close? Don't know if it's a problem on my end or yours since others seem to be able to access it, however I did just DL another stacked TIF.



#18 F.Meiresonne

F.Meiresonne

    Cosmos

  • *****
  • Posts: 7,802
  • Joined: 22 Dec 2003
  • Loc: Eeklo,Belgium

Posted 14 October 2021 - 01:20 AM

I had issues in downloading it too,; all my Windows computers seem to think it is a txt file. Via Edge i could eventually download it cause it proposed the name + .txt, changing that eventually solved it

The Elf told me that was probably the problem, he was right. Still doing thesame from dropbox does not has that problem...

Guess i will have to 'bind' the .tiff extension somehow to a browser or something, sigh....



#19 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 14 October 2021 - 01:29 AM

Tim,

 

my browser does the same but it actually downloads to the usual folder. Just look in there.


  • Tim J Fowler likes this

#20 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 14 October 2021 - 01:31 AM

In most browser a context menu (right mouse button) should be available when the pointer is over the link. It should have a "save link as" option.

 

Added the data to my download page and added the "download" tag in the HTML code that is supposed to force any HTML5 browser NOT to open the file but download it.

https://www.elf-of-l.../Downloads.html

It is the last entry in the first section "One Shot Color Data".

 

Tim, does that work for you? Please note: very likely the browser will not show a file save dialog but just download.


Edited by the Elf, 14 October 2021 - 02:08 AM.

  • Tim J Fowler likes this

#21 Mike in Rancho

Mike in Rancho

    Apollo

  • -----
  • Posts: 1,487
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 14 October 2021 - 02:48 AM

Awfully nice data, Elf!  Thanks for the link.  This is my first run at it, but saved it to my Tutorial folder (i.e. CN data that I don't soon delete) for further practice later.  bow.gif

 

I first looked for it on your Elf website, after some noted here a possible snag in the link, but it wasn't there yet.  So I clicked here and it just popped up a download dialogue and I said save.  It was a tiff.

 

A bit sedate compared to some of mine, except for the bright blue star.  tongue2.gif

 

I did crop in on all the sides to start off, as it was a bit louder and noisier over there.  Sensor edge glow from shooting on a warm night?

 

The background colored noise was there at the outset, but pretty much went away with the right stretch and whatever else ST does to track stuff like that.  At the very end things were quite nice so I backed off the usual denoise and only ran it pretty lightly.

 

gallery_345094_15786_205575.jpg


  • ntph, the Elf and Tim J Fowler like this

#22 Tim J Fowler

Tim J Fowler

    Viking 1

  • *****
  • Posts: 579
  • Joined: 21 Jul 2020
  • Loc: Seneca, South Carolina

Posted 14 October 2021 - 04:05 AM

In most browser a context menu (right mouse button) should be available when the pointer is over the link. It should have a "save link as" option.

 

Added the data to my download page and added the "download" tag in the HTML code that is supposed to force any HTML5 browser NOT to open the file but download it.

https://www.elf-of-l.../Downloads.html

It is the last entry in the first section "One Shot Color Data".

 

Tim, does that work for you? Please note: very likely the browser will not show a file save dialog but just download.

Thanks, Elf, I'll look when I get home this evening. I'm currently on a different PC.



#23 F.Meiresonne

F.Meiresonne

    Cosmos

  • *****
  • Posts: 7,802
  • Joined: 22 Dec 2003
  • Loc: Eeklo,Belgium

Posted 15 October 2021 - 02:03 PM

This thread is making that i get second thoughts. I intended to skip M31 this season...cause actually i had allready a descent image of it last year...after some fiddling. After all this, indeed M31 is not so difficult, imo. I had about 90 minutes of data, did not guide nor dither at the time.

 

Now i see what 6 hours could do....i am reconsidering....

 

And no fiddling with Ha filter, just plain DSLR on the scope, and well, go...makes it all easier...


  • Tim J Fowler likes this

#24 Tim J Fowler

Tim J Fowler

    Viking 1

  • *****
  • Posts: 579
  • Joined: 21 Jul 2020
  • Loc: Seneca, South Carolina

Posted 15 October 2021 - 07:37 PM

I finally got a chance to work on your stack, Elf. No surprise that your data is top notch. I went for a less colorful approach, but I definitely see the appeal of the disco version smile.gif . I may give that a go when I get some more free time. This version was all done in Siril and GIMP.

Makes me want to get started on my M31, if my free nights and weather (with no Moon) will ever get together. 

Attached Thumbnails

  • M31_120x3min_OSC Siril GIMP small.jpg

Edited by Tim J Fowler, 15 October 2021 - 07:41 PM.

  • F.Meiresonne, the Elf and galacticinsomnia like this

#25 the Elf

the Elf

    Soyuz

  • *****
  • topic starter
  • Posts: 3,859
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 16 October 2021 - 07:49 AM

Tim,

 

really like your processing. All the time I tend to produce dim images with little color assuming this is "natural". Then I compare my result to the colorful versions posted on astrobin and take a second attempt with way more color and a brighter over all look. That you call it a "disco version" makes me think I should back up and stay with my initial less saturated style. Really don't know which way to go.

The first time I really added saturation to an image was my recent Cocoon Nebula.

https://www.elf-of-l...Cocoon2021.html

Despite my confession to less saturated images I have to admit I like the candy version. The feedback was positive so I dared using more color again for Andromeda. Ego mode on: in the first place the only person who should like my images is myself. I'm not doing this to please others or as a fishing for compliments thing. Ego mode off. *look to the floor* ... but I also like if others like my images.

One of the best things on CN is Goofi's Imaging Challange. I'm glad nikm continues this great idea. It is absolutely amazing how many different versions of the same object are available in some of the challenges. I only wish the were open for a longer period of time. A single month for me is a time span that often does not offer a single clear night.


  • Tim J Fowler 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