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

Image processing process help. First night imaging.

  • Please log in to reply
20 replies to this topic

#1 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 24 November 2020 - 11:32 PM

Im curious what everyone uses / does for image processing. And what everyone recommends.


I currently have deep sky stacker and Photoshop, I have some custom archsin curves for level adjustments that seem to work well for stretching.


(I've only had one night of imaging as I got the scope very recently, and I just wanted to play around that night and forgot to take calibration frames. So I have no flats/darks/bias)

The processes I have tried so far.

Put lights in DSS
Stack
Adjust the color sliders so that the colors line up and increase saturation
Export to photoshop
Use curve arcsin100 to stretch
Adjust levels
Select stars and increase saturation
Repeat curves adjustment to stretch
Creat a layer that is the color of the sky that should be black and isn't. Then subtract from main layer
Final levels adjustment.

Tonight I saw some different processing advise so I thought I'd try it.

Stack in DSS but don't adjust color or saturation.
Export to photoshop and do everything the same except remove light pollution by using the curve setting that lets you select something that should be black instead of subtracting a layer that is the color of the light pollution.

#2 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 24 November 2020 - 11:33 PM

This was the first edit of orion using the first process I described



#3 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 24 November 2020 - 11:34 PM

This was the first edit of orion using the first process I described

Attached Thumbnails

  • PSX_20201120_213113.jpg


#4 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 24 November 2020 - 11:39 PM

This is the second edit if orion using the second process. 

 

 

Both of these are using the exact same light frames.

 

And sadly I didn't get to get the best set of frames. I only got 10-15 frames at 45sec.  

 

 

Also I noticed in both images that the hot pixels got stacked creating lines of hot pixels.  I thought those where removed automatically?

At least photoshop normally does that. 

 

 

 

Is there any preprocessing I should do before stacking? Other than calibration frames which I will be doing in the future.

I have seen some people will calibrate and do other things to each light frame before stacking. How does that work? 

Attached Thumbnails

  • PSX_20201124_233246.jpg


#5 SkyHunter1

SkyHunter1

    Ranger 4

  • *****
  • Posts: 348
  • Joined: 09 Nov 2017
  • Loc: Long Island NY

Posted 24 November 2020 - 11:56 PM

I use levels, selective color, vibrance, brightness/contrast, mostly. Get annies astro tools, hasta la vista green and astronomy tools for some good AP actions as well. I don't find myself using curves too often. I use only photoshop and plug ins.

 

See my gallery for examples of processing in my signature below.

 

Regards,

Skyhunter1



#6 jrcrilly

jrcrilly

    Refractor wienie no more

  • *****
  • Posts: 35,906
  • Joined: 30 Apr 2003
  • Loc: NE Ohio

Posted 24 November 2020 - 11:57 PM

You must calibrate before aligning and stacking. A median combine will get rid of the warm pixel trails (average or sum combination will keep them). The sky isn't black; if you set the sky to black you are losing data. Best to avoid fiddling with saturation or  color balance until you get the stacked image into Photoshop or whatever you use.


  • SkyHunter1 likes this

#7 bobzeq25

bobzeq25

    ISS

  • *****
  • Posts: 22,909
  • Joined: 27 Oct 2014

Posted 25 November 2020 - 12:12 AM

Very decent start.  Some suggestions below.

 

Most important thing first.  Do not process your data in DSS.  It does a decent job of calibrating/stacking, a _really_ lousy job of processing, including stretching.  It's all too easy for DSS to stretch your data, you need to be sure it doesn't.   You'll know you did it correctly if the image is _really_ dark when you open it in Photoshop.

 

Many people use the DSS/Photoshop combination.  For a beginner I like Astro Pixel Processor.  There are real advantages to an astro specific program that both stacks and processes, and APP is a good one.  Has an excellent tool for reducing the effects of light pollution.  Easier to learn than the games you have to play with a terrestrial editing program like Photoshop.  Not free, but definitely "worth it".

 

You need to start shooting the calibration frames; bias, flats, darks, right away.  The minor point is that they'll improve your images.  The major point is that, otherwise you're very likely to learn bad habits in processing.  Processing is hard enough without having to unlearn bad habits.

 

You don't want the background to be pure black, dark gray is more realistic, and you'll lose some dim detail by making the background "too" black.

 

All this will be much clearer and simpler with APP.  Honest.  <smile>

 

I use PixInsight, an excellent program that most people find hard to learn.  Astro specific, stacks and processes, good tools for reducing the effects of light pollution.  You could think of APP as "PixInsight Lite", does a good job, much easier to learn.  Time spent with APP will not be wasted if you later go to PixInsight, the workflow is similar, the user interface in APP _much_ more intuitive.


Edited by bobzeq25, 25 November 2020 - 12:17 AM.


#8 SkyHunter1

SkyHunter1

    Ranger 4

  • *****
  • Posts: 348
  • Joined: 09 Nov 2017
  • Loc: Long Island NY

Posted 25 November 2020 - 12:19 AM

You must calibrate before aligning and stacking. A median combine will get rid of the warm pixel trails (average or sum combination will keep them). The sky isn't black; if you set the sky to black you are losing data. Best to avoid fiddling with saturation or  color balance until you get the stacked image into Photoshop or whatever you use.

Forgot to mention this as well when writing... I find the above to be exactly right. My advice is don't use DSS to stretch or modify the stack. Do that in photoshop. As advised above and below, I was obsessed with pure black backgrounds until i learned over time to avoid this practice...

 

One caveat to that though... I HATE how my images look on a smartphone! You'll find that's how you show them 99% of the time. if the background isn't pure black it tends to look washed out. I usually have 2 versions because of this.

 

Regards,

Skyhunter1


Edited by SkyHunter1, 25 November 2020 - 12:26 AM.


#9 17.5Dob

17.5Dob

    Fly Me to the Moon

  • ****-
  • Posts: 7,084
  • Joined: 21 Mar 2013
  • Loc: Colorado,USA

Posted 25 November 2020 - 12:20 AM

#1...Do NOT do any adjustments in DSS...

#2 The sky is not black

#3 Never use levels, curves are your friend.



I have migrated to PixInsight..aka PI


This was my first M42 from my dark site processed in PS

26757791729_7315aff0c4_c.jpg

This is 3 years later, using the same original data, but restacked /reprocessed in PI, just a few days ago,

50634659506_e80f976e7f_c.jpg
 


  • elmiko, limeyx, jonnybravo0311 and 1 other like this

#10 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 25 November 2020 - 11:58 AM

You must calibrate before aligning and stacking. A median combine will get rid of the warm pixel trails (average or sum combination will keep them). The sky isn't black; if you set the sky to black you are losing data. Best to avoid fiddling with saturation or  color balance until you get the stacked image into Photoshop or whatever you use.

I appreciate all the advice everyone is giving me. I noticed a few people agreed with jrcrilly that I should calibrate before aligning and stacking. 

 

What does this mean? And how do you do it. 

 

Thanks



#11 jrcrilly

jrcrilly

    Refractor wienie no more

  • *****
  • Posts: 35,906
  • Joined: 30 Apr 2003
  • Loc: NE Ohio

Posted 25 November 2020 - 12:19 PM

calibrate before aligning and stacking. 

 

What does this mean? And how do you do it. 

 

Thanks

At the very least, it means to apply darks before alignment. After alignment, the darks will be useless because the affected pixels will be in the wrong place. May as well do the flats as well, although the small movements caused by alignment probably won't have any measurable effect - but why go through multiple calibrations?.



#12 bobzeq25

bobzeq25

    ISS

  • *****
  • Posts: 22,909
  • Joined: 27 Oct 2014

Posted 25 November 2020 - 12:41 PM

I appreciate all the advice everyone is giving me. I noticed a few people agreed with jrcrilly that I should calibrate before aligning and stacking. 

 

What does this mean? And how do you do it. 

 

Thanks

APP will make the answer to that _completely_ obvious.  As well as answering many other questions.

 

If you don't already know, you'll want to come here to find out how to do bias (a very few cameras require dark flats instead, it's unlikely), flats, and darks.  They're what you calibrate with.  You need all three.  APP plainly shows you how to use them.  You will not need to read any documentation or tutorials.   You don't need to understand the math, APP handles it for you.

 

I _don't_ recommend APP because I think it will make better images for you, although it will. 

 

I recommend it because it will make you a better imager.  My advice to beginners is always aimed at helping them to become the best imager they can be.

 

If you're not interested in that, or can't afford the minimal cost for APP; and just want to mess around, ignore me.  <smile>


Edited by bobzeq25, 25 November 2020 - 12:50 PM.


#13 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 25 November 2020 - 01:50 PM

APP will make the answer to that _completely_ obvious.  As well as answering many other questions.

 

If you don't already know, you'll want to come here to find out how to do bias (a very few cameras require dark flats instead, it's unlikely), flats, and darks.  They're what you calibrate with.  You need all three.  APP plainly shows you how to use them.  You will not need to read any documentation or tutorials.   You don't need to understand the math, APP handles it for you.

 

I _don't_ recommend APP because I think it will make better images for you, although it will. 

 

I recommend it because it will make you a better imager.  My advice to beginners is always aimed at helping them to become the best imager they can be.

 

If you're not interested in that, or can't afford the minimal cost for APP; and just want to mess around, ignore me.  <smile>

 

I will likely purchase either APP or PI eventually. But not for a while. I appreciate the advice though.



#14 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 25 November 2020 - 01:53 PM

I whent back and restacked using median. And I was able to get rid of the hot pixels. 

 

I noticed that all my lights load in as gray 16 bit. 

 

I am using a cannon dlsr, shooting in raw. 

 

Do I need to convert these to tif before stacking? And if so is there an easy way to this? Thanks

Attached Thumbnails

  • PSX_20201125_134531.jpg

  • limeyx likes this

#15 17.5Dob

17.5Dob

    Fly Me to the Moon

  • ****-
  • Posts: 7,084
  • Joined: 21 Mar 2013
  • Loc: Colorado,USA

Posted 27 November 2020 - 08:46 PM

I whent back and restacked using median. And I was able to get rid of the hot pixels. 

 

I noticed that all my lights load in as gray 16 bit. 

 

I am using a cannon dlsr, shooting in raw. 

 

Do I need to convert these to tif before stacking? And if so is there an easy way to this? Thanks

They are "grey", until they have been de-bayered, which occurs further down the processing chain once you hit "OK"


  • bobzeq25 likes this

#16 jonnybravo0311

jonnybravo0311

    Viking 1

  • -----
  • Posts: 617
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 28 November 2020 - 12:45 AM

You are clipping the shadows far too much. Don't worry - we've ALL done it. Space really isn't that black :). You've got maybe 11 minutes of integration time, so the results are going to be noisy and messy. That'll be compounded by the fact you took no calibration frames at all. I guess what I'm saying is not to expect miracles from the minimal data you've obtained.

 

However, what you have done is a great start!

 

On a completely separate note, Galactic Beard Productions and the accompanying logo? Very well done.



#17 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 02 December 2020 - 03:17 PM

You are clipping the shadows far too much. Don't worry - we've ALL done it. Space really isn't that black smile.gif. You've got maybe 11 minutes of integration time, so the results are going to be noisy and messy. That'll be compounded by the fact you took no calibration frames at all. I guess what I'm saying is not to expect miracles from the minimal data you've obtained.

 

However, what you have done is a great start!

 

On a completely separate note, Galactic Beard Productions and the accompanying logo? Very well done.

 

Thank you very much. A friend of mine does graphic design for a living and he helped me out with the logos. 



#18 bhood17

bhood17

    Explorer 1

  • -----
  • topic starter
  • Posts: 52
  • Joined: 09 Jun 2020

Posted 02 December 2020 - 03:21 PM

They are "grey", until they have been de-bayered, which occurs further down the processing chain once you hit "OK"

Can you explain this in more detail please I'm still confused. 

 

 

 

Also I have since taken more images of orion to try to see how more time and calibration does for it.  I got 2hours of exposure, 40 minutes of darks, 40 flats, 40 bias. I noticed that the noise when zooming into image is significantly better now.  I did have troubles with the flats tho. They would reverse my vengiting and add a white glow from the corners.  I took my flats by putting a white shirt over the scope and pointing at a light on my house. Then set the camera to AV mode. 



#19 jonnybravo0311

jonnybravo0311

    Viking 1

  • -----
  • Posts: 617
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted 02 December 2020 - 05:55 PM

Can you explain this in more detail please I'm still confused. 

 

 

 

Also I have since taken more images of orion to try to see how more time and calibration does for it.  I got 2hours of exposure, 40 minutes of darks, 40 flats, 40 bias. I noticed that the noise when zooming into image is significantly better now.  I did have troubles with the flats tho. They would reverse my vengiting and add a white glow from the corners.  I took my flats by putting a white shirt over the scope and pointing at a light on my house. Then set the camera to AV mode. 

Hmmm... let's see if I can explain it.

 

Your camera's sensor is just a bunch of tiny little photoelectric cells. Their job is to turn photons into electrons. The amount of photons the sensors collect determines how strong the electrical signal is. In front of the sensor is a pattern of colors: red, green and blue. Typically, this pattern is arranged as RGGB. The raw file from your camera really doesn't know anything about color (which is why you can get away with setting your white balance to something completely bizarre, but the image can still be processed just fine). All that file contains is effectively "At this point in time, photosite x had an electrical signal strength of y".

 

That's why the raw images look monochrome - they haven't had the color pattern (that RGGB) applied yet. That's what debayering does. It constructs a color image from the raw data based on the mosaic pattern in front of the sensor.

 

The pre-processing of AP data doesn't actually do the debayering until later on in the flow.

 

Hope that helps :)



#20 fewayne

fewayne

    Apollo

  • *****
  • Posts: 1,134
  • Joined: 10 Sep 2017
  • Loc: Madison, WI, USA

Posted 03 December 2020 - 10:28 AM

You are clearly asking the right sorts of questions. And this is a good place to get specific questions answered. But you will save yourself a lot of time if you spend the time and a small amount of money to get one of the books on the topic. I am a big fan of Charles Bracken's The Deep-Sky Imaging Primer (as everyone on this forum knows!), but there are some other good ones too, such as Steve Richard's Making Every Photon Count and a whole bookshelf-full of good ones from our own Jerry Lodriguss.

 

Any of these will give you a solid grounding in not just the what, but the why.



#21 jerr

jerr

    Ranger 4

  • -----
  • Posts: 387
  • Joined: 03 Feb 2008
  • Loc: Czarny Las, Poland

Posted 03 December 2020 - 10:48 AM

Hi Bhood,

 

I have already posted it in another thread, but anyhow I'm posting the below - so sorry for repeating myself if you’ve read that.

 

I've recalled a great tutorial dedicated to processing of M42 done years ago by Anna Morris. Here it is:

 

https://www.epriseph...of-orion-nebula

Enjoy!




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