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

Learning how to take flats with SGPro and G3 16200

  • Please log in to reply
6 replies to this topic

#1 Manav

Manav

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 20 Sep 2016
  • Loc: Sydney, NSW, Australia

Posted 07 September 2019 - 08:27 AM

So I decided to listen to some advice and take some flats for calibration.

 

Initial Setup

 

1. Used a light tracing box with a stepless dimmer similar to this -> https://www.amazon.c...373661262&psc=1

2. Added couple of tracing sheets to achieve diffusion

3. Placed it on top of scope and kicked of flats calibration tool in SGPro and here is the result. https://flic.kr/p/2hbQXh1

4. I combined all the 25 x flats to create a master flat in nebulosity as follows - https://drive.google...rnkuln-_4eW5iAT

 

Observation

 

I have no idea if the master flat (i.e. file linked in point 4 above) is correct?

 

Secondly, by adding this in nebulosity to the light frame makes everything look worst than before  -> https://flic.kr/p/2hbS5ZE

 

What did I do wrong? Any suggestions are welcome.

 

Regards

 

 



#2 jdupton

jdupton

    Surveyor 1

  • *****
  • Posts: 1938
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 07 September 2019 - 08:56 AM

Manav,

 

   The flat you posted looks normal to me. However, the Light frame you calibrated with it looks like the flat over-corrected. (The flat has dark dust spots and a dark top due to (expected) vignetting but the calibrated light frame shows white dust spots and a bright top. Both are due to over-correction by the flat.)

 

   Did you also take any Flat-Dark frames? You need to use some Dark Frames of the same exposure length as the Flat frames. You would then subtract the Flat-Dark frames from the Flat frame before using it to calibrate your target light frames.

 

   In addition, you say you subtracted the flat frame from the light frame. I think you mean you let Nebulosity use the flat frame to calibrate the target light frames. Technically, the actual process that Nebulosity will use internally is one of division rather than subtraction but in any case, the calibrated light frame shows that Nebulosity did the calibration properly.

 

   Tray taking some Flat-Dark frames now and calibrate your Flat. Then run that calibrated flat through Nebulosity and see if you get better results. You should.

 

 

John


Edited by jdupton, 07 September 2019 - 08:58 AM.

  • Manav likes this

#3 Manav

Manav

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 20 Sep 2016
  • Loc: Sydney, NSW, Australia

Posted 08 September 2019 - 06:15 AM

Hi John - First of all I processed flat data at a later date to the light data (using lum filter) and I didnt change the focus in anyway when I moved the scope indoors.

 

I then decided to use the nebulosity pre-processing guide from here -  https://www.google.c...vtW3jmeaa1FLlD2

 

The steps I took in nebulosity were as follows:

 

1. I took 25 flats using lum filter and stacked them togather i.e. align and combine.

2. I took 14 darks using the same time and temperature to what the light data was taken at and then stacked them togather i.e. align and combine

3. I then selected batch and pre-process option where I selected my master dark, master flat and applied them to each image file. i.e. I had 8 lum frames to which I applied master dark and master flats.

4. I then normalised intensities for each of these 8 processed files (normalise light levels)

5. I then batch demosic and square raw color to all 8 processed files (No idea why! I'm just following instructions)

6. I then took the 8 proccessed files and aligned them using align and combine translation method (but save each file not stack).

7. I then for the final time (again) align and combine but this time use the stacking function std. dev filter (1.5 - typical). (Again no idea why)

8. I look the at the final image and cry.....bawling.gif

 

 


   Did you also take any Flat-Dark frames? You need to use some Dark Frames of the same exposure length as the Flat frames. You would then subtract the Flat-Dark frames from the Flat frame before using it to calibrate your target light frames.

No idea I had to do this. frown.gif

 

Next Steps

 

I'm going to do some more reading and testing and update this thread once completed. (Its been a tough weekend)


Edited by Manav, 08 September 2019 - 06:21 AM.


#4 jdupton

jdupton

    Surveyor 1

  • *****
  • Posts: 1938
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 08 September 2019 - 08:21 AM

Manav,

 

   Other than the one missing step, your pre-processing operations look correct. You just need to add a couple of steps. The steps in red font should be added as shown below.

  • Take 25 Flat Frames (Use the same temperature as your Target Light Frames.) (You are doing this now as Step 1 in your post above.)
     
  • Take ~15 dark frames of the same exposure length as your Flat Frames. Stack these shorter Dark Frames to make a Master Flat-Dark Frame.

    (These are not the same as the Dark Frames you took before. You need two sets of Dark Frames for calibration -- one set matches the exposure of the Flat Frames and the other set matches the exposure of your Target's Light Frames.)
     
  • Calibrate your Flat Frames using this Master Flat-Dark Frame. (The Flat-Dark is subtracted from each Flat Frame you took.) (There is no need to try to align Darks or Flats.)
     
  • Now Stack all of the Flat Frames that have had the Master Flat-Dark Frame subtracted. This becomes your Master Flat Frame.
     
  • Take ~14 (14 is OK) Dark Frames of the same exposure and temperature as the Target Light Frames. Stack these to make your Master Dark Frame. (This is your Step 2 above.)
     
  • At this point, pick up at your step 3 as you outlined above and finish your normal processing.

   I think once you have calibrated your Flat Frames with a Master Flat-Dark Frame, your image calibration will look much better. You can try to take the Flat-Dark frames now. Since they are dark, you only need to match the temperature at which you took the original flat frames and use the same exposure you used for the flat frames you already have. Then retry the processing but add the steps for calibrating the Flat Frames with the new Master Flat-Dark frame. Use that new calibrated Master Flat Frame along with your Master Dark Frame (not the Master Flat-Dark Frame) to calibrate all of your Target's Light Frames.

 

 

John


  • Manav likes this

#5 namsel

namsel

    Vostok 1

  • *****
  • Posts: 120
  • Joined: 21 Nov 2013
  • Loc: Cambridge, MA

Posted 15 September 2019 - 04:17 PM

John - those are some helpful awnsers to understanding how to process flats, darks and lights, etc.

 

Do you have any links you could post to understand better the processing of darks and flats, generally, and how to best come up for a workflow for most use-cases ?

 

Namsel



#6 jdupton

jdupton

    Surveyor 1

  • *****
  • Posts: 1938
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 15 September 2019 - 04:46 PM

Namsel,

 

   The processing is highly dependent on which software you are using to process images. I would start by suggesting that you do a Google search for
"<insert_the_name_of_your_software_of_choice_here> preprocessing tutorial"

such as "Deepskystacker preprocessing tutorial" or "pixinsight preprocessing tutorial".

 

   Two suggestions will give you a good introduction if using PixInisght are: David Ault's Preprocessing Tutorial and Chis Foster's image processing work-flow. David Ault's methods are for using CCD cameras while Chis' write-up is more specific to CMOS cameras. Both are helpful once you understand the basics involved. The best source for getting the big picture and undertsnading the basics is probably a book. Either of the book by Bracken or Woodhouse will give you a great leg up in seeing how all aspects fit together.

 

   In addition, there a a great number of processing tutorials on YouTube.

 

 

John


  • Manav likes this

#7 namsel

namsel

    Vostok 1

  • *****
  • Posts: 120
  • Joined: 21 Nov 2013
  • Loc: Cambridge, MA

Posted 16 September 2019 - 05:54 PM

Beautiful, thank you so much John, for the guidance.




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