Jump to content


Photo

Processing help

  • Please log in to reply
24 replies to this topic

#1 woolbrig

woolbrig

    Mariner 2

  • -----
  • Posts: 295
  • Joined: 01 Feb 2013
  • Loc: Bethalto, Illinois

Posted 28 June 2013 - 03:08 PM

Last night was my first chance to shoot after fixing most of my auto guiding issues, so this is my 1st image guiding. It's 6-300 sec subs and 17 darks.

I'm seeing a lot of black spots in the image and don't know what I'm doing to cause them or what I need to do to get rid of them. Any ideas? This is a 100% crop

Attached Files



#2 CounterWeight

CounterWeight

    Cosmos

  • *****
  • Posts: 8160
  • Joined: 05 Oct 2008
  • Loc: Palo alto, CA.

Posted 28 June 2013 - 04:05 PM

What sort of stacking did you do? did you try more than one type? Average, median, sum, sigma ? I used to get something like that when I used an OSC CCD and all I could think of was something to do with noise calibration and temps as it was an unregulated CCD.

#3 woolbrig

woolbrig

    Mariner 2

  • -----
  • Posts: 295
  • Joined: 01 Feb 2013
  • Loc: Bethalto, Illinois

Posted 28 June 2013 - 05:02 PM

I stacked them in Nebulosity 3.
I tried Average and Std. Dev filter 1.5

The pic above was done doing Average. I'm a newbie to using Nebulosity too.

The temp for the lights was 97 and the darks were 102 is that to far apart?

#4 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6333
  • Joined: 14 Jan 2013

Posted 28 June 2013 - 05:44 PM

I have gotten those kinds of results in Nebulosity too. It use to drive me crazy. There are probably ways of dealing with it in Nebulosity, but I switched to other processing (PixInsight).

#5 Ron (Lubbock)

Ron (Lubbock)

    Messenger

  • -----
  • Posts: 438
  • Joined: 17 Aug 2012
  • Loc: West TX

Posted 28 June 2013 - 05:46 PM

I have gotten black dots like that before with my CCD, but never so many. I agree that a median filter or sigma-kappa clipping could help remove them, but I don't use the same software you use. Also check to see if there is an option to automatically detect and remove hot and cold pixels while stacking.

#6 CounterWeight

CounterWeight

    Cosmos

  • *****
  • Posts: 8160
  • Joined: 05 Oct 2008
  • Loc: Palo alto, CA.

Posted 28 June 2013 - 05:48 PM

For a CCD I would say yes that it maybe too hot, but since you are using a DSLR I hope someone else can help. Glad to hear you tried different types, that is always my first impulse as each has things it might be better or at least different depending on what you are after.

#7 woolbrig

woolbrig

    Mariner 2

  • -----
  • Posts: 295
  • Joined: 01 Feb 2013
  • Loc: Bethalto, Illinois

Posted 28 June 2013 - 09:56 PM

I guess I'll try again in Deep Sky Stacker.

#8 *skyguy*

*skyguy*

    Surveyor 1

  • *****
  • Posts: 1982
  • Joined: 31 Dec 2008
  • Loc: Western New York

Posted 28 June 2013 - 10:18 PM

Looks like bad dark frames. Are they old, taken during a previous imaging session and/or at a different duration or temperature than the light frames? Try Re-shooting the darks and see if it makes a difference. Good Luck ...

#9 LoveChina61

LoveChina61

    Viking 1

  • *****
  • Posts: 924
  • Joined: 19 Jun 2009

Posted 29 June 2013 - 12:41 AM

the rule I try to follow is to take new Darks for every 2 degree up/down change in temperature. It seems that I read somewhere to do that a few years ago and I have kept to that rule ever since.

Anyone else follow a different rule on that?

#10 Raginar

Raginar

    Fly Me to the Moon

  • *****
  • Posts: 6138
  • Joined: 19 Oct 2010
  • Loc: Rapid CIty, SD

Posted 29 June 2013 - 11:17 AM

China,

I got a regulated CCD and don't worry about it. I take new darks every quarter or so just to make sure the data is fresh.

#11 woolbrig

woolbrig

    Mariner 2

  • -----
  • Posts: 295
  • Joined: 01 Feb 2013
  • Loc: Bethalto, Illinois

Posted 29 June 2013 - 02:19 PM

The darks I took were taken the same night as the lights.

I took some more darks last night. I'll try to match the temperature of the darks to the lights and see if that helps.

#12 Raginar

Raginar

    Fly Me to the Moon

  • *****
  • Posts: 6138
  • Joined: 19 Oct 2010
  • Loc: Rapid CIty, SD

Posted 29 June 2013 - 10:49 PM

It'll make a huge difference if the temperature deviation was significant. Check your flats too. I see some pretty weird things when my flats suck.

Oh, and try taking like 200 bias frames. I know it sounds stupid, but if you google in the PI forums... you'll see the reason why. It's pretty ridiculous.

#13 RedLionNJ

RedLionNJ

    Viking 1

  • -----
  • Posts: 893
  • Joined: 29 Dec 2009
  • Loc: Red Lion, NJ, USA

Posted 30 June 2013 - 11:01 AM

With a DSLR, I'd speculate the temperatures are a little high to hope to get good results (did you say 300s exposures?) without a full (large) of good calibration frames. If the camera sensor wasn't at close to equilibrium, your recorded temperatures for darks (and lights) could just be somewhat-educated guesses.

The symptom is definitely indicative of the darks being at significantly-higher temperatures than the lights. As a previous poster suggested, you really need the darks to be within a couple degrees of the lights (and definitely not hotter) and to take as many darks, flats and bias (or flat-dark) frames as you can. What seem like ridiculously high numbers do seem to make a difference, esp in summer.

Personally, I find I'm not up to the challenge of creating really good DSLR long exposures in the NJ summer temperatures. My DSLRs don't see action like this until the night-time temperatures fall into the sixties.

I also didn't see mention of an ISO. It's possible the use of too high an ISO is just multiplying the "darker" noise and making it seem even more exaggerated after basic post-processing.

Grant

#14 jwaldo

jwaldo

    Smart Mime

  • *****
  • Posts: 3923
  • Joined: 26 Apr 2004
  • Loc: Simi Valley, CA

Posted 30 June 2013 - 01:03 PM

This looks like the same issue I struggled with in my M51 this week, though to a greater degree. Temperature-related dark frame trouble is my guess too. It's been HOT here (80F at midnight), my camera was hot to the touch by the end of the session. I'm absolutely sure my light and dark temperatures don't match. I'm still trying to solve the issue, but I might end up having to wait for it to cool off a bit...

#15 SergeC

SergeC

    Apollo

  • *****
  • Posts: 1125
  • Joined: 23 Mar 2012
  • Loc: Gainesville, FL

Posted 30 June 2013 - 02:30 PM

Try this with Nebulosity:
A. Stack your darks:
1. Go to Batch->Align and Combine Images
2. In Output Mode click Save stack
2. Click None (fixed) under Alignment method
3. Click Average/Default under Stacking function
4. Uncheck Fine-tune star location
5. Select your dark frames to stack
6. Save the fit file when prompted

If you collected dark flats and bias frames, repeat the above steps for those frames and name them appropriately.

B. Make a bad pixel map from your master dark stack:
1. Go to Batch->Make Bad Pixel Map
2. Select the master dark stack you saved above
3. You'll see a number of white pixels and some darker grey pixels on a light grey background. Adjust the slider to increase the number of black pixels so that all of the white pixels and most of the darker grey pixels become black.
4. Save this bad pixel map as a fits file.

C. Pre-process the light frames.
1. Go to Batch->Pre-process image sets
2. Click Dark 1 and select the bad pixel map file. From the drop-down list select BPM RAW color.
3. If you're using dark flats load the averaged master you made in step A above as Dark 2 and leave the drop-down selection as Dark subtract.
4. Load the master bias as Bias 1 (if using bias)
5. Load the master flat as Flat 1. If using dark flat, do not subtract bias, choose Dark 2, and set 7 pixel blur.
6. Load your light frames in Light 1, select Bias 1, Dark 1, and Flat 1.
7. Leave stack method as Average, and the file prefix as pproc_ and hit OK.

D. Debayer the pproc_files using Batch->Batch Demosaic+RAW Color.

E. File->Preview to load the recon_pproc_ files examine each one. Delete any you don't want to stack (e.g. bad tracking). Don't worry about satellite trails, they'll disappear in the next step.

F. Align the recon_ files - SAVE A COPY OF EACH ALIGNED FILE.
1. Batch->Align and Combine Images
2. In output mode click Save each file
3. In Alignment method choose appropriate method
4. Click OK and go through the star selection process to align the images. When finished selecting, Nebulosity will create new files with the prefix align_

G. Final stacking of aligned and processed light frames:
1. Batch->Align and Combine Images
2. In output mode click Save stack
3. In alignment method click None (fixed)
4. In stacking function click one of the Std. Dev. filter options. Usually 1.5 or 1.75 is sufficient, but for noisy light frames try 2.0 or even custom and enter a larger number. The higher the SD, though, the more detail you'll lose.

#16 woolbrig

woolbrig

    Mariner 2

  • -----
  • Posts: 295
  • Joined: 01 Feb 2013
  • Loc: Bethalto, Illinois

Posted 30 June 2013 - 08:50 PM

Wow! Thanks for the step by step processing. I followed it using the new darks I collected and was able to improve my initial image. Here it is at 100% crop

Attached Files



#17 woolbrig

woolbrig

    Mariner 2

  • -----
  • Posts: 295
  • Joined: 01 Feb 2013
  • Loc: Bethalto, Illinois

Posted 30 June 2013 - 08:53 PM

Here is the full picture. I still need to work on my guiding and processing, but I think it turned out ok for my first guiding attempt.

The colors in the jpg look a little crushed compared to the tiff.

Attached Files



#18 jwaldo

jwaldo

    Smart Mime

  • *****
  • Posts: 3923
  • Joined: 26 Apr 2004
  • Loc: Simi Valley, CA

Posted 30 June 2013 - 10:48 PM

Wow, that's an improvement! :jump:

Now I need to figure out how to do all that in Deep Sky Stacker :tonofbricks:

#19 JoseBorrero

JoseBorrero

    Gemini

  • *****
  • Posts: 3144
  • Joined: 04 Sep 2009
  • Loc: Puerto Rico Island

Posted 01 July 2013 - 01:55 AM

oh yeah! great job! :waytogo:

#20 CounterWeight

CounterWeight

    Cosmos

  • *****
  • Posts: 8160
  • Joined: 05 Oct 2008
  • Loc: Palo alto, CA.

Posted 01 July 2013 - 03:58 AM

Looks like something worked :) ! That's turning into a fine image!

#21 woolbrig

woolbrig

    Mariner 2

  • -----
  • Posts: 295
  • Joined: 01 Feb 2013
  • Loc: Bethalto, Illinois

Posted 01 July 2013 - 07:56 AM

Well, after looking at the image for a while, I decided I wasn't happy with it so I gave it another go at processing. I think this one looks better. If you have suggestions on how to improve it, let me know.

Also forgot to mention the specs for the shot earlier.

12 - 300sec subs (over 2 nights)
15 - Darks
17 - Flats
10 - Dark Flats
50 - Bias

Software: BackyardEOS, Nebulosity, Photoshop CS3, Astronomy Tools, GradientXTerminator

Attached Files



#22 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6333
  • Joined: 14 Jan 2013

Posted 01 July 2013 - 09:02 AM

I agree this new version is much better. :)

#23 LoveChina61

LoveChina61

    Viking 1

  • *****
  • Posts: 924
  • Joined: 19 Jun 2009

Posted 01 July 2013 - 09:09 AM

Yep. It keeps getting better and better. The last picture is about as good a picture as anyone could possibly take of M27. Great job!

#24 Greyhaven

Greyhaven

    Surveyor 1

  • *****
  • Moderators
  • Posts: 1856
  • Joined: 11 May 2004
  • Loc: 16 miles south of the Buckfield Mall

Posted 01 July 2013 - 11:09 AM

Great reprocessing, I think that you've created a real work of art vibrant colors and tons of details. :bow:
Be Well
Grey

#25 RedLionNJ

RedLionNJ

    Viking 1

  • -----
  • Posts: 893
  • Joined: 29 Dec 2009
  • Loc: Red Lion, NJ, USA

Posted 01 July 2013 - 05:06 PM

Looks significantly better. Really good job! My personal preference is for a background sky somewhere between the shade of your last two (i.e. a tad darker than the last one, but not jet black). But that's just MY personal preference when processing.

Hope you have many more clear skies!

Grant






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics