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

Yes, another M31 -- What can I Improve?

  • Please log in to reply
21 replies to this topic

#1 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 28 September 2020 - 08:08 PM

I am up to maybe a dozen evenings out, only a small number of them where the clouds let me get much, only two where I made it to morning. I am even further from understanding enough of Pixinsight.

But I'm starting to get things that look like what I expect them to.  cool.gif
 
This was taken 9/22 and 9/26, with 332 total exposures of 60 seconds each, with an unmodified Sony A7Riv and a 400/2.8 lens with a 1.4x teleconverter (making it F4).  I know teleconverter use is rare but this framed it nicely and gave me 1.3" per pixel which is probably optimistic that I got anywhere near good enough seeing to use it, but hope springs eternal.  It was guided with a 60mm guide scope, and captured with N.I.N.A and processed in Pixinsight with touchup in Photoshop. 
 
Any and all criticism solicited -- too saturated, not enough, color, stars (too many, too few -- I reduced them a bit).  More subs, shorter, longer.  Whatever. 

 

M31_PI_Final.jpg
 
This is a full resolution copy: https://www.captivep...31_PI_Final.jpg

 

Thanks in advance for any criticism.

 

Linwood


  • F.Meiresonne, mikewayne3, mrlovt and 6 others like this

#2 Bob Koch

Bob Koch

    Lift Off

  • -----
  • Posts: 4
  • Joined: 05 Oct 2013

Posted 28 September 2020 - 08:15 PM

I think this is one of the more interesting renditions of M31. Subtle coloration and highlights. Nicely done!

 

  Bob


  • Linwood likes this

#3 17.5Dob

17.5Dob

    Fly Me to the Moon

  • ****-
  • Posts: 6,897
  • Joined: 21 Mar 2013
  • Loc: Colorado,USA

Posted 28 September 2020 - 08:16 PM

A "bit" over saturated..but clicking the link.......you have some really serious CA on all of your stars.....ditch the TC and stop down your lens...


Edited by 17.5Dob, 28 September 2020 - 08:25 PM.

  • limeyx and Linwood like this

#4 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 28 September 2020 - 08:17 PM

A "bit" over saturated..but you have some really serious CA on all of your stars.....

Yeah, I'm actually searching for whether I could have addressed that in post.  I think it's partly from use of a teleconverter.  So I'm thinking, if I can't fix it well, the TC might stay in the bag next time.

 

May I ask whether you meant "bit" literally or did the quotes mean it is grossly over-saturated?  (I'm too new at this to read between the lines). 



#5 imtl

imtl

    Apollo

  • *****
  • Moderators
  • Posts: 1,029
  • Joined: 07 Jun 2016

Posted 28 September 2020 - 08:18 PM

Hi Linwood,

 

In my opinion you are off to a very good start.

 

If you want some input then I see a few things.

 

1. You got some chromatic aberrations. You can see your stars are half blue half orage. Its probably the lens you are using. It can be dealt with in PI though.

2. You got some aperture vignetting but again that is probably due to your optical system and you're trying to do F/4.

3. The image is over saturated to my taste. It causes a flare up of your stars. If you feel that you want to emphasize the galaxy's color more then protect your stars with masking.

4. You should run SCNR to reduce the blue overcast you got.

 

Good start!


  • Linwood likes this

#6 DuPont

DuPont

    Mariner 2

  • -----
  • Posts: 231
  • Joined: 19 Mar 2020
  • Loc: Fond du Lac, Wi.

Posted 28 September 2020 - 08:20 PM

Very nice image Linwood. The detail & color of M31 is spot on in my opinion. I've been imaging M31 for the last week but unfortunately there's been smoke to deal with which is really screwing with my outcome.

The stars look tight but maybe a little to red, but other then that I would be very happy if I could get your results........Al.


  • Linwood likes this

#7 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 28 September 2020 - 08:46 PM

1. You got some chromatic aberrations. You can see your stars are half blue half orage. Its probably the lens you are using. It can be dealt with in PI though.

So I cheated a bit and toned down the CA in Photoshop. It's not gone but better (not sure you can see it at this size).  I need to read more on how to deal with it in PI, but I also need to do a bit of testing to see how much is coming from the 1.4x TC.  I'm guessing that may be the cause of much of it. 
 

2. You got some aperture vignetting but again that is probably due to your optical system and you're trying to do F/4.

Now that's interesting. I actually put a gradient on the left two and bottom right corner in Photoshop to darken them a bit, as I had some red brightness there that I thought looked bad.  I really thought I had matched it up well.  Thank you.

 

I'm not sure what you mean by "trying to do F4" -- meaning I should have stuck to F2.8?  Or that F4 was fast and part of the issue? 
 

4. You should run SCNR to reduce the blue overcast you got.

Thank you.  I actually added adjusted the green that was there to be mostly gone but brought up the blue on purpose as I had seen shots of M31 that had a blue outer tone.

 

That brings up a question: I realize for nebula with sharp spectral spikes this question is not exactly answerable, but this is a galaxy -- is there a "correct" color, i.e. how a human would see it at a suitable distance for the same field of view by eye? 

 

Sorry I forgot to attach the new version: 

 

M31_PI_Final.jpg

 

The link above is updated, though if you already look be sure to refresh.


Edited by Linwood, 28 September 2020 - 08:48 PM.

  • George Simon likes this

#8 imtl

imtl

    Apollo

  • *****
  • Moderators
  • Posts: 1,029
  • Joined: 07 Jun 2016

Posted 28 September 2020 - 09:19 PM

So I cheated a bit and toned down the CA in Photoshop. It's not gone but better (not sure you can see it at this size).  I need to read more on how to deal with it in PI, but I also need to do a bit of testing to see how much is coming from the 1.4x TC.  I'm guessing that may be the cause of much of it. 

 

Probably so. In PI when you align the RGB channels you can choose distortion correction and that usually takes care of stuff like this. I get these CAs when I shoot lower altitude objects that yields a significant atmospheric dispersion.
 

Now that's interesting. I actually put a gradient on the left two and bottom right corner in Photoshop to darken them a bit, as I had some red brightness there that I thought looked bad.  I really thought I had matched it up well.  Thank you.

 

I'm not sure what you mean by "trying to do F4" -- meaning I should have stuck to F2.8?  Or that F4 was fast and part of the issue? 

 

No I did not mean vignetting in that sense. I mean some of your bright stars have this black spikes in them. This is usually from either pinched optics or from aperture vignetting. Don't worry about it there is very little you can do about it at this stage. You can try and process it out but I wouldn't worry about it now.
 

Thank you.  I actually added adjusted the green that was there to be mostly gone but brought up the blue on purpose as I had seen shots of M31 that had a blue outer tone.

 

Its a matter of taste of course.

 

That brings up a question: I realize for nebula with sharp spectral spikes this question is not exactly answerable, but this is a galaxy -- is there a "correct" color, i.e. how a human would see it at a suitable distance for the same field of view by eye? 

 

There is no "correct" way in what we are doing. Its all manipulation of data and perception. Its a mix of science and art. The right way is the one your eyes like the most. Don't worry about a standard way. There is none. Humans do have a specific color perception of course. Mostly green and red. Not a lot of blue. What you would "see" with your eyes when looking up is a washout white blob :) Unless you start going to some serious aperture and get enough photons to excite the cones in your eyes. The rods are "color blind". 

 


  • dswtan and Linwood like this

#9 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 28 September 2020 - 09:24 PM

No I did not mean vignetting in that sense. I mean some of your bright stars have this black spikes in them. This is usually from either pinched optics or from aperture vignetting. Don't worry about it there is very little you can do about it at this stage. You can try and process it out but I wouldn't worry about it now.

 

Ah... yes, it does that. I don't know of anything I can do. Think of it as inverse diffraction spikes. smile.gif


  • imtl likes this

#10 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 28 September 2020 - 09:31 PM

 

Probably so. In PI when you align the RGB channels you can choose distortion correction and that usually takes care of stuff like this. I get these CAs when I shoot lower altitude objects that yields a significant atmospheric dispersion.

I've got another set for M8 (etc) I'm just starting with the same lens set, so almost certainly the same issue.  Thanks for the pointer, I'll give it a try.

 

But my guess is he real answer may be to take off the TC.  I was hoping the 1.4x on an already fast lens was a decent compromise to get a bit more focal length, but I suspect it's the cause and not using it may be cleaner.

 

Thank you!


Edited by Linwood, 28 September 2020 - 09:31 PM.


#11 imtl

imtl

    Apollo

  • *****
  • Moderators
  • Posts: 1,029
  • Joined: 07 Jun 2016

Posted 28 September 2020 - 09:32 PM

Ah... yes, it does that. I don't know of anything I can do. Think of it as inverse diffraction spikes. smile.gif

If it makes you happy then sure :p :p :p



#12 imtl

imtl

    Apollo

  • *****
  • Moderators
  • Posts: 1,029
  • Joined: 07 Jun 2016

Posted 28 September 2020 - 09:33 PM

I've got another set for M8 (etc) I'm just starting with the same lens set, so almost certainly the same issue.  Thanks for the pointer, I'll give it a try.

 

But my guess is he real answer may be to take off the TC.  I was hoping the 1.4x on an already fast lens was a decent compromise to get a bit more focal length, but I suspect it's the cause and not using it may be cleaner.

 

Thank you!

Too much non corrected optics will hurt you. Take it out.



#13 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 28 September 2020 - 09:41 PM

Too much non corrected optics will hurt you. Take it out.

Yeah, I'm finding the KISS principle is pretty important here.  Well, except when it comes to post processing.


  • F.Meiresonne likes this

#14 HoloTheWolf

HoloTheWolf

    Sputnik

  • *****
  • Posts: 31
  • Joined: 22 Jul 2017
  • Loc: Northern California

Posted 29 September 2020 - 01:17 AM

Nice work! It captures many details of the spiral disk.



#15 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 29 September 2020 - 11:27 AM

Probably so. In PI when you align the RGB channels you can choose distortion correction and that usually takes care of stuff like this. I get these CAs when I shoot lower altitude objects that yields a significant atmospheric dispersion.

This is a OSC, so I'm not aligning R, G, and B separately.  The process I am following is to calibrate, then debayer, then star align.

 

I tried using the Thin plate Splines with Distortion Correction and Local Distortion and compared to the default options running through integration with each.

 

Interestingly off in the corners I found some double stars that became single -- I THINK when it did a median flip there was enough distortion in the corners that the undistorted did not actually align, the distortion version fixed that.

 

But I didn't see any reduction in CA.  Here's both integrations side by side with saturation boosted. 

 

Are you aligning R, G, and B separately in some way, maybe extracting the separate channels first then combining after alignment? 

 

ca.jpg



#16 imtl

imtl

    Apollo

  • *****
  • Moderators
  • Posts: 1,029
  • Joined: 07 Jun 2016

Posted 29 September 2020 - 11:38 AM

I use mono so I have separate channels. You can separate the channels with an OSC as well I guess.

Might be better to remove the TC as mentioned above instead of post processing.



#17 Palmito

Palmito

    Explorer 1

  • -----
  • Posts: 86
  • Joined: 16 Nov 2018
  • Loc: Lausanne, Switzerland

Posted 29 September 2020 - 11:39 AM

This is a OSC, so I'm not aligning R, G, and B separately.  The process I am following is to calibrate, then debayer, then star align.

 

I tried using the Thin plate Splines with Distortion Correction and Local Distortion and compared to the default options running through integration with each.

 

Interestingly off in the corners I found some double stars that became single -- I THINK when it did a median flip there was enough distortion in the corners that the undistorted did not actually align, the distortion version fixed that.

 

But I didn't see any reduction in CA.  Here's both integrations side by side with saturation boosted. 

 

Are you aligning R, G, and B separately in some way, maybe extracting the separate channels first then combining after alignment? 

 

attachicon.gifca.jpg

I have been suffering from the same CA when I was shooting with a camera lens (100-400 F/5.6). I was using a mono astro camera with filters.

 

I found no way of fixing it. Camera lenses are not apochromatic and are not optimized for infinity focus.

 

Those issues disappeared when I switched to an apochromatic refractor. I am afraid you will have to live with those CA or change OTA :-/



#18 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 29 September 2020 - 06:49 PM

Well, so far attempts to manage the CA have been unsuccessful.

 

One suggestion is that if caused by refraction in the atmosphere, the colors will be on the same side on all portions of the image, and you can shift colors and recombine -- that's not the case here, it's opposite sides.

 

Another suggested splitting the integrated image, star aligning, and recombining.  I tried it with the stretched image without any success -- it wouldn't star align at all. With the one fresh out of integrate it would align, but actually looked worse rather than better.  I think it's because of shape -- the red portion on the left (for example) is like have a star flaring left.  It doesn't really match the other half flaring right, except generally at the bright center, which is roughly where it was before.

 

I'm curious now that I think about it why the pixinsight process calls for debayering before star align.  I wonder if it wouldn't be better to align in the CFA matrix then debayer.  But of course it doesn't.  I could separate the CFA layers and align each of those to each other, but not sure the eventual debayer would work (greens would be on top of each other rather than beside). 

 

Or debayer and then split and align, getting more like an RGB workflow, but I am not sure how that is different from doing it on the integrated (other than taking a hugely longer time). 

 

So... a work in progress.

 

I do think the local distortion helped with alignment at the corners where after median flip they did not align well, so I'll add that to my process.



#19 JamesTX

JamesTX

    Mariner 2

  • -----
  • Posts: 262
  • Joined: 27 Aug 2017
  • Loc: Texas

Posted 29 September 2020 - 07:28 PM

Hope you dont mind.. I downloaded your pick and tried a couple things.

 

For the CA.. there is a package available for photoshop called astronomy tools.  One of the "actions"/scripts is for reducing blue/violet halos.  I ran that against your picture and it did a real good job with the CA.

 

Next a loaded it up in pixinsight and did the old invert>scnr>invert trick to remove the magenta (mostly from the stars).  Normally on M31 you would want to mask the stars before hitting it with scnr otherwise it'll kill the Ha in M31.  Here's what I got (see attached).

 

I'd get more data.  :)

 

 

 

 

Attached Thumbnails

  • m31-mod.JPG

  • Palmito, limeyx and Linwood like this

#20 F.Meiresonne

F.Meiresonne

    Fly Me to the Moon

  • *****
  • Posts: 6,373
  • Joined: 22 Dec 2003
  • Loc: Eeklo,Belgium

Posted 30 September 2020 - 02:11 AM

Nice image. Bit too blue  in the spiral arms as other mentioned but anyhow a nice image.Just my opinion.



#21 WoodlandsAstronomer

WoodlandsAstronomer

    Vostok 1

  • -----
  • Posts: 148
  • Joined: 26 Aug 2019
  • Loc: Texas

Posted 30 September 2020 - 02:25 AM

I like it, I assume you did a composite with that nice tight and just right exposure of the core? Colors strike me as slightly off, but not much. Well done and keep at it :)

Edited by WoodlandsAstronomer, 30 September 2020 - 02:26 AM.


#22 Linwood

Linwood

    Viking 1

  • -----
  • topic starter
  • Posts: 836
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 30 September 2020 - 10:50 AM

Hope you dont mind.. I downloaded your pick and tried a couple things.

 

For the CA.. there is a package available for photoshop called astronomy tools.  One of the "actions"/scripts is for reducing blue/violet halos.  I ran that against your picture and it did a real good job with the CA.

 

Next a loaded it up in pixinsight and did the old invert>scnr>invert trick to remove the magenta (mostly from the stars).  Normally on M31 you would want to mask the stars before hitting it with scnr otherwise it'll kill the Ha in M31.  Here's what I got (see attached).

 

Don't mind at all, appreciate it.  I'll look for that package; I'm curious if it does more than the layer duplicate/blur/color blend.

 

I'm off processing another set of data now, and trying to use some of what I learned, then will come back to my M31.  I still haven't found much to address CA early in the process (at least that is working).  It seems wrong to let it all integrate wrong, spread left and right, then fix it after.  Of course the real solution is better optics.  Getting rid of the teleconverter.

 

 

I'd get more data.  smile.gif

Didn't you read the terms and conditions here?  Every posting is presumed to end with the phrase "you need more integration time".   waytogo.gif




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