Problem with calibr...
 
Share:
Notifications
Clear all

Problem with calibration frames  

Page 1 / 3
  RSS

(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 22, 2020 18:48  

Hi there. 

After more then 6 hours of stacking,i get my result:

NGC7023 session 1 session 2 session 3 session 4 session 5 session 6 St

I was expecting the result, because I got a message in the beginning about not fitting Calibration Frames. The data is collected from 6 sessions, and I suspect something is wrong with one of them. Is there any way I can use APP to quickly find the source of the error in the data? If necessary I have to skip a session.

Thanks for help!


ReplyQuote
(@wvreeven)
Black Hole Admin
Joined: 2 years ago
Posts: 482
November 22, 2020 20:40  

Hi Dane,

In general the following rules apply:

  • All frames (light, bias, dark, flat and darkflat) should be taken at the same ISO/gain and temperature. Flat and darkflat may be different than the rest but should be the same as each other.
  • Darks should have the same exposure time as lights.
  • Darkflat should have the same exposure time as the flats.
  • This applies to all frames within a session but this may differ between sessions.
  • If calibration frames (bias, dark, flat, darkflat) are used between sessions then the rules above apply again for those sessions.

Can you verify that this is the case for all frames in all 6 sessions?

 

Wouter


ReplyQuote
(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 23, 2020 18:22  

Hi Wouter

  • All frames (light, bias, dark, flat and darkflat) should be taken at the same ISO/gain and temperature. Flat and darkflat may be different than the rest but should be the same as each other.
    I have a ASI 2600 MC Pro, and i use only gain 100 for all exposures. I never change this setting. So here im save.
  • Darks should have the same exposure time as lights.
    I have a dark library with 3, 5 and 10minutes exposures per -10 and -20 degree. But on this case its simple, used for every shot a 10 minute exposure at -10 degree. So i have not the problem with diffrent exposures in this case. Here im also save.
  • Darkflat should have the same exposure time as the flats.
    I do flats and darkflats with the same settings at the same time. first with the flat light on the scope, than with the cap on the scope. On both situations i use the same settings. So that should be also save.

     

  • This applies to all frames within a session but this may differ between sessions.
    I know that is use a shorter exposure time (100ms) on my first and second session for flats and flatdarks. On the later sessions i use 2 seconds. So its diffrent between the sessions.But the frames within the sessions should be similar.
  • If calibration frames (bias, dark, flat, darkflat) are used between sessions then the rules above apply again for those sessions.
    I do for every session flats, darkflats. I don't use any bias when I do darkflats, so far I had no problems.  And the darks i use the same for alle sessions - 10min at -10 degrees. 

I use the Asiar as computer, and the put the exposure time, gain, binning into the filename. So i can check fast througt aall folders and i can see easy if there something with other settings. But i found nothing. On every session, flats and dark flats has the same exposure time. Every light has the same exposure time. All frames have the same gain and temperature. Temperature not perfect alway on -10, sometimes on -9.3 to -10.6. But i think that should be not the problem for die bright vignette?

Thats so strange, i cant found any mistake. Or its maybe really the small fluctuation in temperature?

EDIT: I take a look into the stacking folder, and i see this files. 

2020 11 23 18 48 38 Window

There is a file missing for session 3. So the problem muste be in session 3? MDF = Master Dark File? 

I see in session 3, i have 30 flats, 30 dark flats, but only 15 lights. Can this be the problem das i have more calibration frames as light frames there?

Thank you!

This post was modified 2 months ago 5 times by Dane Vetter

ReplyQuote
(@wvreeven)
Black Hole Admin
Joined: 2 years ago
Posts: 482
November 23, 2020 20:11  

Hi Dane,

As it happens, I do part of my imaging with an ASI2600MC on gain 100 and temperature -10º as well. I use KStars on a raspberry pi 4 using the latest drivers of ZWO so that's pretty similar to your setup. One thing I learned with ZWO hardware (I own an ASI1600MM and an ASI6200MM as well) is to avoid exposure times shorter than about 1 second. The CMOS chips used by ZWO simply don't perform well enough to give reliable calibration frames. As a check, could you only stack those sessions that have flats and darkflats with exposure times of 2 seconds and see if that solves the issue? If it does, we can discuss the way you shoot your flats to see how the exposure time can be increased for those of 200 ms.

 

HTH, Wouter


ReplyQuote
(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 23, 2020 20:20  

Hi Wouter,

yes it works fine with short exposures. But i want to go save and use 2 seconds to be sure to get an homogeneous flat. At the first and second session, i used the short exposure times bei 100ms. Here is an stack from the first session. So the calibration frames with short times should be not the problem. 

NGC7023 RGB session 1 lpc cbg St

But to return to my previous question. In session 3 there i no MDF, so i guess at this session i must get the error message. 

The lights of session 3: 

image

The flats of session 3:

image

The dark flats of session 3:

image

Maybe im blind and i not see a bad frame?


ReplyQuote
(@wvreeven)
Black Hole Admin
Joined: 2 years ago
Posts: 482
November 24, 2020 10:17  

Hi Dane,

I tag @mabula-admin so he can reply to this thread.

 

Wouter


ReplyQuote
(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 24, 2020 21:04  

I do some tests, and I still do not understand the problem!

As mentioned above I suspected that the problem was in session 3. So I placed all sessions except session 3, and see: It worked without problems! So here my assumption was confirmed.

Now for test, i simply stacked only session 3 alone. And see there: I can stack the data without any problems.

So I tried again to stack all 6 sessions. And what happens, I get the error message again! This makes me crazy and is totally illogical what happens here.

image

ReplyQuote
(@wvreeven)
Black Hole Admin
Joined: 2 years ago
Posts: 482
November 24, 2020 21:59  

Hi Dane,

Sorry, I am a bit lost in all your info here. Does session 3 share any data with other sessions, notably sessions 1 and 2? If yes, can you try to stack only those sessions (including session 3) and verify that that reproduces the error that you get? If you then leave out session 3, does the error not happen? If all of this is confirmed then we may be getting closer to understanding what is going wrong.

 

Thanks, Wouter


ReplyQuote
(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 24, 2020 23:04  

Hi Wouter,

i try to do this the next days. Now i startet a new round to stack all 6 sessions with the same files like on my first trys. Now i get another error message like before. But on this message, i can only press on ok, and its goes further. 

2020 11 24 22 46 44 Astro Pixel Processor version 1.082

Perhaps it would be useful to be able to classify these error messages more precisely. If you have more than one session you can't identify which session has the problem. Or an even better idea: The message directly states which flats do not harmonize with which dark flats oder bias. The program detects the error, so it must be able to tell where exactly it does not fit. 


ReplyQuote
(@wvreeven)
Black Hole Admin
Joined: 2 years ago
Posts: 482
November 25, 2020 00:35  

Hi Dane,

The popup you now get is not an error but a warning. It indicates what the issue is though it could indeed be improved. I’m tagging @mabula-admin to see what he can say about that proposal.

Wouter


ReplyQuote
(@vincent-mod)
Universe Admin
Joined: 4 years ago
Posts: 2885
November 25, 2020 16:02  

That warning actually means that flats are not going to be used (effectively), reason being they were, apparently, not corrected with a proper bias/darkflat. The issue is probably at that stage, what then the exact issue is we would probably need to see for ourselves with actual data.


ReplyQuote
(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 25, 2020 17:10  

Hello everyone. I could not believe my eyes this morning, but the stack was successful! No idea why it worked this time. And that despite the warning.

Now I have two more questions about the completed stack. I hope you would have some hints.

1. the dark clouds do not stand out so clearly from the background. I've seen shots with a fraction of the exposure time with identical equipment where the dark clouds stand out better. Are there any settings that I could maybe optimize to make them stand out better?

2. there is still a clear noise, especially in the dark cloud on the right. Maybe there are more artifacts. Is there a setting that could possibly improve this?

NGC7023 session 1 session 2 session 3 session 4 session 5 session 6 lpc cbg csc St klein

Thank you!


ReplyQuote
(@wvreeven)
Black Hole Admin
Joined: 2 years ago
Posts: 482
November 25, 2020 21:04  

Hi Dane, great that your stack looks good now. Regarding your first question: now post-processing begins and you’ll need to play with the colors and contrast. Regarding your second question: there is no noise reduction in APP yet but this is planned for a future version. 

Wouter


ReplyQuote
(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 25, 2020 21:07  

Hi Wouter,

thank you. So i hope for more settings to prevent or reduce noise during the stacking process. 


ReplyQuote
(@wvreeven)
Black Hole Admin
Joined: 2 years ago
Posts: 482
November 25, 2020 21:32  

Hi Dane,

Just a small piece of info: noise reduction typically is a process applied during post-processing and not during stacking. 

Wouter


ReplyQuote
(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 25, 2020 21:39  

Hi Wouter,

and what is with Dithering between the Shots? And Darks are also to reduce the noise during the stacking? Or am I wrong? Post processing would be much better if you already had good source material from the stacking, right? Maybe I have a wrong way of thinking about this. 🙂


ReplyQuote
(@wvreeven)
Black Hole Admin
Joined: 2 years ago
Posts: 482
November 25, 2020 22:55  

@dv_stranger Hi Dane,

Yes you are right. Dithering and using darks help reduce noise. However, there are more ways of noise reduction which can be user controlled in post-processing and that will come to APP in a future version.

 

Wouter


ReplyQuote
(@vincent-mod)
Universe Admin
Joined: 4 years ago
Posts: 2885
November 26, 2020 00:17  

Well, slight correction/additive. Darks can actually introduce some noise, dithering is a good way to get rid of a number of fixed pattern noise indeed and darks for amp-glow and hot pixels mainly.


ReplyQuote
(@mabula-admin)
Universe Admin
Joined: 4 years ago
Posts: 2507
November 26, 2020 14:00  
Posted by: @dv_stranger

Hi Wouter,

i try to do this the next days. Now i startet a new round to stack all 6 sessions with the same files like on my first trys. Now i get another error message like before. But on this message, i can only press on ok, and its goes further. 

2020 11 24 22 46 44 Astro Pixel Processor version 1.082

Perhaps it would be useful to be able to classify these error messages more precisely. If you have more than one session you can't identify which session has the problem. Or an even better idea: The message directly states which flats do not harmonize with which dark flats oder bias. The program detects the error, so it must be able to tell where exactly it does not fit. 

Hi Dane, @dv_stranger

I am glad that you were able to get a good integration in the end. I think your issue is caused by overlap of some of the darkflats in multiple sessions:

If you assign a batch of darkflats for session 3 first, then assign the same batch for session4 or session5, then the frames are no longer assigned to session3, therefore the flats of session3 are not calibrated, leading to the warning message and subsequent error in the stack.

My advice is to NOT assign the darkflats to separate sessions. I would simply assign them to all sessions. If you have darkflats of different exposure times (and different gains) then APP will automatically create different MasterDarkFlats anyway and will match them correctly with the flats. It makes things a bit more easy that way 😉

Mabula


ReplyQuote
(@dv_stranger)
Main Sequence Star Customer
Joined: 1 year ago
Posts: 41
November 26, 2020 14:23  

Thats an very important information! 

I use two types of darks. I renewed my dark library between sessions. That means in session 1-3 I use Darks A, and session 4-6 I use Darks B. So I can't use the function "assign to all sessions" here at all. That means I insert Darks A in session 1, then I insert Darks A in session 2, and then I insert Darks A in session 3.

But in reality I have darks only in session 3, and in the previous sessions the darks were deleted because they are exactly the same? The same for sessions 4-6, then I only have darks B in session 6 and not in 4, 5.

But: I have for evere session it own dark flats, so i assign for every session it own flats. But i can use alway the same dark flats when the exposure time is the same like the flats? No matter from which session?


ReplyQuote
Page 1 / 3
Share: