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

Please help me find what's wrong with my flat frame

Software Imaging DSO Astrophotography
  • Please log in to reply
14 replies to this topic

#1 maxhy

maxhy

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 11 Nov 2020
  • Loc: Hangzhou, China

Posted 17 May 2021 - 10:29 AM

This is my first time processing deep sky images with flat frames and I am using DeepSkyStacker to stack my light frames and flat frames of Leo Triplet.

But the result image looks very different from what I wanted to get and it seems that flat frames make vignetting even worse and flat frames erased some details from light frames (lightframe-only stack on the right side):

 

With flat frames:                                                                                                                                               Without flat frames

 

with-flatframes.png               without-flats.png

 

I used my 16-inch Macbook Pro screen with a pure white background to capture my flat frames (by holding it as close as can to the telescope).

I took flat frames with same focus/temp/gain right after my light frame image session (exposure time 1.4s and 31 flat frames).

I adjusted exposure time several times to move most of the tops to 1/3~1/2 of the histogram graph and below is how my flat's histogram looks:

Flatframe-histogram.png

 

Here are the stacking params I used:

Stacking-params.png

 

Could some one help me identify the problem? Thanks in advance!

 

And could some one tell me why my stacked picture is greenish? I consulted ZWO tech support about the debayer algorithm I should use and I used RGGB in DSS FITS file settings as they suggested.

If I am using the correct debayer algorithm, the resulting color should be more "natural" instead of this weird green, I guess?

 

 

Equipments I used:

ASI533 MC Pro

HEQ5pro

SkyWatcher 150 PDS


Edited by maxhy, 17 May 2021 - 10:30 AM.


#2 NuTek

NuTek

    Sputnik

  • *****
  • Posts: 41
  • Joined: 16 Jun 2016

Posted 17 May 2021 - 10:44 AM

Following as I have the same issues (same camera)



#3 DubbelDerp

DubbelDerp

    Surveyor 1

  • *****
  • Posts: 1,700
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 17 May 2021 - 10:49 AM

Calibration with flats will not work correctly unless you calibrate your master flat with either bias or dark flats of the same ISO/gain as the flats. Shoot some bias or dark flats and it should work much better. Good luck!


  • jdupton likes this

#4 maxhy

maxhy

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 11 Nov 2020
  • Loc: Hangzhou, China

Posted 17 May 2021 - 11:07 AM

Calibration with flats will not work correctly unless you calibrate your master flat with either bias or dark flats of the same ISO/gain as the flats. Shoot some bias or dark flats and it should work much better. Good luck!

By bias you mean just take ordinary bias frame with the fastest shutter speed ?



#5 adosaj

adosaj

    Explorer 1

  • *****
  • Posts: 93
  • Joined: 04 Jul 2017

Posted 17 May 2021 - 11:08 AM

It is likely that the light source you are using does not have a flat spectral response. It is very important to invest in a light source that puts out light evenly across the visible spectrum. Even if it looks white to your eyes, it like is not even across all wavelengths. 
 

Anil


  • maxhy likes this

#6 DubbelDerp

DubbelDerp

    Surveyor 1

  • *****
  • Posts: 1,700
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 17 May 2021 - 11:14 AM

By bias you mean just take ordinary bias frame with the fastest shutter speed ?

Correct. Temperature is not significant, but fastest exposure time at same ISO/gain as your flats and no light hitting the sensor. Or dark flats, which is same ISO/gain as your flats, AND same exposure time. Pick one or the other, not both.


  • maxhy likes this

#7 maxhy

maxhy

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 11 Nov 2020
  • Loc: Hangzhou, China

Posted 17 May 2021 - 11:27 AM

It is likely that the light source you are using does not have a flat spectral response. It is very important to invest in a light source that puts out light evenly across the visible spectrum. Even if it looks white to your eyes, it like is not even across all wavelengths. 
 

Anil

I am planning to invest a light panel. But can you tell from the histogram I provided to make sure this is the problem? 

Flatframe-histogram.png



#8 maxhy

maxhy

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 11 Nov 2020
  • Loc: Hangzhou, China

Posted 17 May 2021 - 11:28 AM

Correct. Temperature is not significant, but fastest exposure time at same ISO/gain as your flats and no light hitting the sensor. Or dark flats, which is same ISO/gain as your flats, AND same exposure time. Pick one or the other, not both.

OK Thanks, I'll take some bias/dark frames later and redo the stacking to see if it'll help!



#9 DubbelDerp

DubbelDerp

    Surveyor 1

  • *****
  • Posts: 1,700
  • Joined: 14 Sep 2018
  • Loc: Upper Peninsula of Michigan

Posted 17 May 2021 - 11:44 AM

Your histogram looks close enough. I use sky flats that are distinctly blue, so some mismatch between the peaks in the color channels will be corrected by your calibrating/stacking program. There are plenty of folks that use inexpensive LCD screens or tablets as their light source, so you should be able to get it to work as long as your exposures are long enough to not catch a "flicker" in some screens. I try to check and fix the simple things first before changing up the gear. The screen could possibly be an issue. Not calibrating your flats, however, always is.


  • maxhy likes this

#10 jdupton

jdupton

    Gemini

  • *****
  • Posts: 3,375
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 17 May 2021 - 11:56 AM

maxhy,

 

   Your light source looks good and the raw Flat frames very likely to be good. None of the color channel peaks are close to either edge. The problem you are having with the Master Flat frame over-correcting leading to reverse vignetting is common. There are multiple other active and recent threads here on CN dealing with this very same subject. 

 

   The issue is most likely due to calibration. You need to use either a Master Bias frame or a Flat-Dark frame to calibrate you individual raw Flat frames. Then you stack those calibrated Flat frames to make a Master Flat. You also want to stack your Dark frames to make a Master Dark. Finally, you will use the Master Dark and the Master Flat to calibrate and then stack your Light frames.

 

   If your Master Flat frame has not been calibrated with either a Master Bias or Master Flat-Dark, then Light calibration will fail to give the results you want.

 

If your Light frames are not then calibrated with the Master Dark and Master Flat, then the Light calibration will fail to give the results you want and the stacked image will show problems like you are experiencing.

 

   See the following active thread for more detail on the process.

 

https://www.cloudynights.com/topic/771544-asi2600mmmc-f63-focal-reducer-vignette-question/?view=findpost&p=11106408

 

   This next post also summarizes the information above.

 

https://www.cloudynights.com/topic/768291-flat-frames-causing-reverse-vignetting/#entry11058877

 

 

John


  • maxhy likes this

#11 maxhy

maxhy

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 11 Nov 2020
  • Loc: Hangzhou, China

Posted 17 May 2021 - 12:04 PM

Your histogram looks close enough. I use sky flats that are distinctly blue, so some mismatch between the peaks in the color channels will be corrected by your calibrating/stacking program. There are plenty of folks that use inexpensive LCD screens or tablets as their light source, so you should be able to get it to work as long as your exposures are long enough to not catch a "flicker" in some screens. I try to check and fix the simple things first before changing up the gear. The screen could possibly be an issue. Not calibrating your flats, however, always is.

Yeah this makes sense, I try calibration first...



#12 maxhy

maxhy

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 11 Nov 2020
  • Loc: Hangzhou, China

Posted 17 May 2021 - 12:06 PM

maxhy,

 

   Your light source looks good and the raw Flat frames very likely to be good. None of the color channel peaks are close to either edge. The problem you are having with the Master Flat frame over-correcting leading to reverse vignetting is common. There are multiple other active and recent threads here on CN dealing with this very same subject. 

 

   The issue is most likely due to calibration. You need to use either a Master Bias frame or a Flat-Dark frame to calibrate you individual raw Flat frames. Then you stack those calibrated Flat frames to make a Master Flat. You also want to stack your Dark frames to make a Master Dark. Finally, you will use the Master Dark and the Master Flat to calibrate and then stack your Light frames.

 

   If your Master Flat frame has not been calibrated with either a Master Bias or Master Flat-Dark, then Light calibration will fail to give the results you want.

 

If your Light frames are not then calibrated with the Master Dark and Master Flat, then the Light calibration will fail to give the results you want and the stacked image will show problems like you are experiencing.

 

   See the following active thread for more detail on the process.

 

https://www.cloudynights.com/topic/771544-asi2600mmmc-f63-focal-reducer-vignette-question/?view=findpost&p=11106408

 

   This next post also summarizes the information above.

 

https://www.cloudynights.com/topic/768291-flat-frames-causing-reverse-vignetting/#entry11058877

 

 

John

Thanks John for detailed reply, I'll try what you and DubbelDarp suggest tommorrow !



#13 adosaj

adosaj

    Explorer 1

  • *****
  • Posts: 93
  • Joined: 04 Jul 2017

Posted 17 May 2021 - 03:35 PM

I guess I never picked up that the individual flat frames were not being corrected for dark current, but if the flat source is, say less bright in the blue as compared to the red, you will not correctly “flatten” your pixel response. I hope I’m explaining this correctly. I don’t understand the yellow’ish hue in the non-flat corrected image, are you using a filter? Also, your histogram should reflect the pixels at about 2/3 full well. 


  • maxhy likes this

#14 jdupton

jdupton

    Gemini

  • *****
  • Posts: 3,375
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 17 May 2021 - 04:03 PM

Anil,

 

I guess I never picked up that the individual flat frames were not being corrected for dark current, but if the flat source is, say less bright in the blue as compared to the red, you will not correctly “flatten” your pixel response. I hope I’m explaining this correctly. I don’t understand the yellow’ish hue in the non-flat corrected image, are you using a filter? Also, your histogram should reflect the pixels at about 2/3 full well. 

   If the color channels don't have the same peak locations in the histogram, the Master Flat frame will still perfectly flatten the Light. However, the relative strength of the channels will be "reversed" in the calibrated histogram.

 

   In other words, if the Blue channel is the weakest in the Master Flat, it will be boosted such that it will be the strongest in the calibrated Light. The calibrated Light frame will still be fully flattened with respect to dust motes and vignetting but the overall color balance will change. For example, if the blue channel was weakest in the Master Flat, the blue channel will become the strongest channel in the calibrated light and still have been fully corrected for "flatness".

 

   That color change is of no consequence, though. After the images are stacked, the next step will be to either extract gradients or neutralize the background. Either operation will help restore the original colors of the image target. You would normally then follow up with a full color calibration step.

 

   If using PixInsight for calibration of the Light frames, there is an option called "Separate CFA Scaling Factors" that defaults to being selected for OSC images. This option treats the OSC image as if it already has three different color channels (even though it is Grey scale at that point) and flattens each channel independently. This will result in no color channel strength changes during the Light calibration process. If blue was the weakest, it will remain the weakest after calibration.

 

   Either method is normally followed by gradient extraction, background neutralization, and color calibration. The only difference is whether the color channels retain their original relative strengths after Light calibration. They are both always flattened so long as the Light frames are being calibrated with a Master Dark and the Master Frame was previously calibrated with either a Master Bias or Master Flat-Dark.

 

 

John


Edited by jdupton, 17 May 2021 - 04:07 PM.

  • adosaj and maxhy like this

#15 adosaj

adosaj

    Explorer 1

  • *****
  • Posts: 93
  • Joined: 04 Jul 2017

Posted 18 May 2021 - 10:43 AM

Anil,

 

   If the color channels don't have the same peak locations in the histogram, the Master Flat frame will still perfectly flatten the Light. However, the relative strength of the channels will be "reversed" in the calibrated histogram.

 

   In other words, if the Blue channel is the weakest in the Master Flat, it will be boosted such that it will be the strongest in the calibrated Light. The calibrated Light frame will still be fully flattened with respect to dust motes and vignetting but the overall color balance will change. For example, if the blue channel was weakest in the Master Flat, the blue channel will become the strongest channel in the calibrated light and still have been fully corrected for "flatness".

 

   That color change is of no consequence, though. After the images are stacked, the next step will be to either extract gradients or neutralize the background. Either operation will help restore the original colors of the image target. You would normally then follow up with a full color calibration step.

 

   If using PixInsight for calibration of the Light frames, there is an option called "Separate CFA Scaling Factors" that defaults to being selected for OSC images. This option treats the OSC image as if it already has three different color channels (even though it is Grey scale at that point) and flattens each channel independently. This will result in no color channel strength changes during the Light calibration process. If blue was the weakest, it will remain the weakest after calibration.

 

   Either method is normally followed by gradient extraction, background neutralization, and color calibration. The only difference is whether the color channels retain their original relative strengths after Light calibration. They are both always flattened so long as the Light frames are being calibrated with a Master Dark and the Master Frame was previously calibrated with either a Master Bias or Master Flat-Dark.

 

 

John

John,

 

Thanks for the great explanation. I only briefly used a DSLR and then moved on to a mono camera. I was extending the definition of flats (as correcting for the pixel to pixel response variation) into the color space, incorrectly attributing the introduction of a color gradient as a “not flatness” in the color space. Of course that was the wrong way to look at this since even in the mono with filters workflow we have to correct for color gradients introduced after combining all the filters. That clears up a misconception on my part :)  Thanks again!

 

Anil




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





Also tagged with one or more of these keywords: Software, Imaging, DSO, Astrophotography



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics