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

Does anyone know what would cause this?

  • Please log in to reply
29 replies to this topic

#1 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 06:56 AM

These are both my master flats for my Antlia 36mm 3.5nm Ha and OIII. I dont get this with my LRGB set. 

In a ZWO 36mm EFW used with a ZWO ASI183MM PRO. Let me know if any other details are needed. 



#2 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 06:57 AM

meant to add them

Attached Thumbnails

  • ANT_HELP.jpg


#3 OhmEye

OhmEye

    Viking 1

  • *****
  • Posts: 821
  • Joined: 15 Sep 2019
  • Loc: Western NY Southern Tier

Posted 01 October 2020 - 07:07 AM

Are they not working for calibration? I'm not sure I see an issue. I see different exposures, but without histogram and details of how you created them or knowing what the problem looks like when you use them, I'm not sure what you are asking.



#4 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 07:11 AM

The weird pattern around the edges. I know its a crappy picture because its a screenshot then changed to JPEG and resized in PAINT. Different exposures, both were around 1/3 of the histogram. I used both a Gerd Neumann flat panel and a basic Amazon tracing pad, both yield this patterning. Its more pronounced in the Ha than the OIII, probably a bit overexposed in the OIII. It gives a weird background pattern in my stacked image. 



#5 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 07:13 AM

Will dive into this a bit more, but my online e.chem class just started



#6 Dynan

Dynan

    Aurora

  • *****
  • Posts: 4,507
  • Joined: 11 Mar 2018
  • Loc: NOLA

Posted 01 October 2020 - 07:24 AM

Upload and link one of each (or their masters) for us to look at if possible.



#7 OhmEye

OhmEye

    Viking 1

  • *****
  • Posts: 821
  • Joined: 15 Sep 2019
  • Loc: Western NY Southern Tier

Posted 01 October 2020 - 07:29 AM

I'm trying to understand if you are troubleshooting an actual problem, or just curious about the way the flats look. I rarely look at my flats that way unless I have an issue using them, I've seen a lot of strange looking flat sets that work perfectly fine for me yet very different from each other, often different filters have very different looking flats, partly due to the wavelengths my light panel has relative to the bandpass of the filter. If you are saying those flats are not working, please show what the calibrated lights look like.

 

Edit: Ahh, I missed where you said it causes a pattern in your stack, sorry. Yes, having the original raw flat would help people make suggestions.


Edited by OhmEye, 01 October 2020 - 07:31 AM.

  • jdupton and wrnchhead like this

#8 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 08:20 AM

Upload and link one of each (or their masters) for us to look at if possible.

I can upload them to a google drive but need an email.



#9 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 08:22 AM

I'm trying to understand if you are troubleshooting an actual problem, or just curious about the way the flats look. I rarely look at my flats that way unless I have an issue using them, I've seen a lot of strange looking flat sets that work perfectly fine for me yet very different from each other, often different filters have very different looking flats, partly due to the wavelengths my light panel has relative to the bandpass of the filter. If you are saying those flats are not working, please show what the calibrated lights look like.

 

Edit: Ahh, I missed where you said it causes a pattern in your stack, sorry. Yes, having the original raw flat would help people make suggestions.

Would a file of like 10 RW FITS help? I cna upload those as well to my drive account but need an email to allow access. 



#10 OhmEye

OhmEye

    Viking 1

  • *****
  • Posts: 821
  • Joined: 15 Sep 2019
  • Loc: Western NY Southern Tier

Posted 01 October 2020 - 08:36 AM

The common method using drive is to just share it public and post the link. That's the simplest way to have people here check out the files if they'd like to help.



#11 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 08:43 AM

oh wasnt aware, can I make just that folder public? Its my school account so i cant make the whole thing public.



#12 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 09:07 AM

Please tell me if this works and also let me know if it is only the flats folder that is viewable to the public. Id rather not have my school drive open to the public jsut for privacy of assignments and presentations. Not that anyone here would compromise my security on there but ya never know...

https://drive.google...2Ow?usp=sharing



#13 jdupton

jdupton

    Mercury-Atlas

  • *****
  • Posts: 2,823
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 01 October 2020 - 09:16 AM

OrionNebula32,

 

   As user OhmEye indicated, the flats for the narrow-band filters you posted look rather ordinary to me. I have also seen a lot of very odd looking flats that work just fine.

 

   You mention that the flats are causing issues that show up in your integrated images. That could be because either one (or both) of the following are true:

  • The issue is showing up in the flats for some reason but not the lights.
    This will cause a pattern to be introduced by the flat-fielding process.
  • or
  • The flats are not being fully calibrated which will result in either over-correction or under-correction. 
    In this case, the pattern does exist in both the lights and the flats but is not being completely cancelled out.

   In order to debug both cases, we would need to look at a sample Light, Flat, Dark, and Flat-Dark (or Bias, whichever you use). I see you have uploaded sample Light and Flats. If you can add a Dark and Flat-Dark or Bias, we can get a better picture of where the issue may lie.

 

EDIT: Oops. I misinterpreted the frames you uploaded. They were all flats. I can use the Ha and Oiii Master Flats but will need a single raw, uncalibrated Light frame for Ha and Oii also. In addition, Your Master Dark and Master Flat-Dark or Master Bias will be needed to see what is going on.

 

 

John


Edited by jdupton, 01 October 2020 - 09:22 AM.


#14 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 09:20 AM

OrionNebula32,

 

   As user OhmEye indicated, the flats for the narrow-band filters you posted look rather ordinary to me. I have also seen a lot of very odd looking flats that work just fine.

 

   You mention that the flats are causing issues that show up in your integrated images. That could be because either one (or both) of the following are true:

  • The issue is showing up in the flats for some reason but not the lights.
    This will cause a pattern to be introduced by the flat-fielding process.
  • or
  • The flats are not being fully calibrated which will result in either over-correction or under-correction. 
    In this case, the pattern does exist in both the lights and the flats but is not being completely cancelled out.

   In order to debug both cases, we would need to look at a sample Light, Flat, Dark, and Flat-Dark (or Bias, whichever you use). I see you have uploaded sample Light and Flats. If you can add a Dark and Flat-Dark or Bias, we can get a better picture of where the issue may lie.

 

 

John

I will upload darks, i dont take bias as the ZWO fix for my 183 amp glow is to not integrate with bias and its worked in the past with just using the camera with no filters(mono)



#15 jdupton

jdupton

    Mercury-Atlas

  • *****
  • Posts: 2,823
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 01 October 2020 - 09:39 AM

OrionNebula32,

 

   OK, I will download the Dark. I think you have confirmed why the Flat pattern is showing up in the stacked images. If the Frame is not also calibrated with either a Flat-Dark or Bias Frame, then the flat-fielding process will not correctly fix the illumination of the image. That will allow the pattern seen in the flats to be imparted onto the Light frames they calibrate.

 

   Do you happen to have any Bias Frames laying around that you could upload for a sample? Even if you do not use them, I can better verify what may be happing if I have one to play with.

 

 

John


  • OrionNebula32 likes this

#16 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 09:45 AM

Have not needed to use bias for any of my other images since getting the filters with my mono. 

I only ever use bias with my 294 for color shots. After my next 2 classes and i have my dog go down for her nap then i will attempt to take some just in case that would be a resolution. Wont be for another few hours tho as i have class until 1:45EST, so 3 hours from now. 



#17 jdupton

jdupton

    Mercury-Atlas

  • *****
  • Posts: 2,823
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 01 October 2020 - 09:47 AM

OrionNebula32,

 

   No hurry on the sample Bias. I can infer a general level by extracting data from the Master Dark Frame in the meantime.

 

 

John


  • OrionNebula32 likes this

#18 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 11:20 AM

Was able to sneak in a few bias between classes, so Im uploading a few of those now



#19 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 01:28 PM

I tried to remove the filter cells and flip them in the EFW slots and get a similar pattern but now its more pronounced on the left side than it is on the right side. I also tried to take a few in SharpCap as well instead of using NINA and its still there. This is very frustrating as I dont make a lot of money and i just spent $1100 on SHO/LRGB filters



#20 jdupton

jdupton

    Mercury-Atlas

  • *****
  • Posts: 2,823
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 01 October 2020 - 02:09 PM

OrionNebula32,

 

   I am having trouble making sense out the files you uploaded. They appear to have different offsets in some cases and some of the capture parameters don't match.

 

   Did you take these using different programs and device drivers? The Master Dark looks suspicious also. Was that taken using live stacking in SharpCap by any chance?

 

   I will keep looking at them but a lot of things don't add up for me at the moment.

 

 

John


  • OrionNebula32 likes this

#21 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 02:44 PM

Yes sorry the darks were taken with SharpCap and rotated 180deg to match my NINA files. Regardless the issue is with my flats. NINA takes images and shows the image taken with autostretch, thats where i get the issue in the first set of images. Those are just RAW FITS files direct from NINA. The darks are with SharpCap. but just rotated and the darks are just the camera noise, nothing to do w the optics. 



#22 jdupton

jdupton

    Mercury-Atlas

  • *****
  • Posts: 2,823
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 01 October 2020 - 05:13 PM

OrionNebula32,

 

   I understand what you are saying. However, I think the root of the issue you are asking about is not with the pattern in the Flat Frames but in a generic calibration problem. Below is a link to a Post I made in another thread where I show two of my Master Flats. One is from my AstroDon LRGB Blue filter and the other is from my AstroDon Ha filter. Their patterns are just as strange as those you posted at the beginning of this thread. 

 

https://www.cloudynights.com/topic/724414-strange-artifacts-with-astrodon-oiii/?view=findpost&p=10436638

 

   My two odd looking filter Flats calibrate out completely and show no residual patterns in my images. I think yours will also fully calibrate out once the calibration methods are modified.

 

   Your description of taking your Flat Frames with SharpCap is the first problem area. You should always use exactly the same software with exactly the same device driver for all frame captures associated with one imaging project. (You can experiment with different software. Just do not mix acquisition of frames that will be used in processing a single image session.)

 

   The reason for not mixing programs or device driver types (ASCOM / Native / SDK) is that each program and each driver can capture and store the frames in slightly different ways. The Dark frame you uploaded showed a different Offset setting was used for the camera than was used for the other frames. The tip off there is that median ADU level of the Dark Frame was 592 ADU. The median of a Bias Frame you supplied was 704 ADU. Since the Dark frame is supposed to also contain the Bias signal inside it, it can never be lower than a matching Bias frame. This type of error is often pretty easy to check. 

 

   Simply look at the statistics for a single sample of each frame type. They should appear in this order of increasing average ADU:

  • Bias Frame -- Lowest average ADU
  • Flat-Dark Frame -- Next lowest average ADU (Not needed if Bias frames available and short Flat exposures.)
  • Dark Frame -- Higher average ADU than the above but lower than other frame types.
  • Light Frame -- Higher average ADU than all previous frame types.
  • Flat Frame -- Highest average ADU of all frame types.

   If your check of the average ADU level of frames shows any out of order when sorted like this, something is not matched and calibration of your images will fail to give a good integrated image.

 

   Your dark frame was shot at a different Offset setting from the others. SharpCap tends to hide or rename parameters so it is easy to get confused. Also, it seems SharpCap may have written out the wrong image ordering causing the darks to appear flipped with respect to your Flats. I did not see a Light so I cannot tell what its bit ordering was.

 

   For calibration to work properly, all parameters used to acquire the images must match one another. The frame samples you uploaded looks like:

  • Flats taken at Gain 111 Offset 10, Temp at -10°C
  • Darks taken at Gain 111, Offset 8 (?), Temp at -10°C
  • Bias taken at Gain 111, Offset 10, Temp +20.5°C
  • Light Frames unknown

   The calibration procedure should consist of:

  • Integrate (Matching) Bias frames to build a Master Bias
  • Integrate (New matching) Dark frames to build a Master Dark.
     
  • Calibrate the Flat Frames with the Master Bias
  • Integrate the calibrated Flat frames to build the Master Flat (for each filter)
     
  • Calibrate the Light Frames with the Master Dark and Master Flat
     
  • Proceed with your normal process flow from this point.

   I think your next steps should be to shoot new Dark Frames using NINA (Gain = 111, Offset = 10, Temp = -10°C). Also shoot new Bias Frames at -10°C using NINA (at Gain = 111, Offset = 10). Follow the preprocessing / Calibration steps outlined above. You should NOT need to modify the new Dark frames in any way at all. If they do not match amp glow patterns with your Light frames, then something else is still going wrong.

 

   After doing the above, I think the pattern may be much more subdued (or hopefully eliminated) in your integrated Light.

 

   Let us know how it goes.

 

 

John


Edited by jdupton, 01 October 2020 - 05:18 PM.

  • OrionNebula32 likes this

#23 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 05:40 PM

OrionNebula32,

 

   I understand what you are saying. However, I think the root of the issue you are asking about is not with the pattern in the Flat Frames but in a generic calibration problem. Below is a link to a Post I made in another thread where I show two of my Master Flats. One is from my AstroDon LRGB Blue filter and the other is from my AstroDon Ha filter. Their patterns are just as strange as those you posted at the beginning of this thread. 

 

https://www.cloudynights.com/topic/724414-strange-artifacts-with-astrodon-oiii/?view=findpost&p=10436638

 

   My two odd looking filter Flats calibrate out completely and show no residual patterns in my images. I think yours will also fully calibrate out once the calibration methods are modified.

 

   Your description of taking your Flat Frames with SharpCap is the first problem area. You should always use exactly the same software with exactly the same device driver for all frame captures associated with one imaging project. (You can experiment with different software. Just do not mix acquisition of frames that will be used in processing a single image session.)

 

   The reason for not mixing programs or device driver types (ASCOM / Native / SDK) is that each program and each driver can capture and store the frames in slightly different ways. The Dark frame you uploaded showed a different Offset setting was used for the camera than was used for the other frames. The tip off there is that median ADU level of the Dark Frame was 592 ADU. The median of a Bias Frame you supplied was 704 ADU. Since the Dark frame is supposed to also contain the Bias signal inside it, it can never be lower than a matching Bias frame. This type of error is often pretty easy to check. 

 

   Simply look at the statistics for a single sample of each frame type. They should appear in this order of increasing average ADU:

  • Bias Frame -- Lowest average ADU
  • Flat-Dark Frame -- Next lowest average ADU (Not needed if Bias frames available and short Flat exposures.)
  • Dark Frame -- Higher average ADU than the above but lower than other frame types.
  • Light Frame -- Higher average ADU than all previous frame types.
  • Flat Frame -- Highest average ADU of all frame types.

   If your check of the average ADU level of frames shows any out of order when sorted like this, something is not matched and calibration of your images will fail to give a good integrated image.

 

   Your dark frame was shot at a different Offset setting from the others. SharpCap tends to hide or rename parameters so it is easy to get confused. Also, it seems SharpCap may have written out the wrong image ordering causing the darks to appear flipped with respect to your Flats. I did not see a Light so I cannot tell what its bit ordering was.

 

   For calibration to work properly, all parameters used to acquire the images must match one another. The frame samples you uploaded looks like:

  • Flats taken at Gain 111 Offset 10, Temp at -10°C
  • Darks taken at Gain 111, Offset 8 (?), Temp at -10°C
  • Bias taken at Gain 111, Offset 10, Temp +20.5°C
  • Light Frames unknown

   The calibration procedure should consist of:

  • Integrate (Matching) Bias frames to build a Master Bias
  • Integrate (New matching) Dark frames to build a Master Dark.
     
  • Calibrate the Flat Frames with the Master Bias
  • Integrate the calibrated Flat frames to build the Master Flat (for each filter)
     
  • Calibrate the Light Frames with the Master Dark and Master Flat
     
  • Proceed with your normal process flow from this point.

   I think your next steps should be to shoot new Dark Frames using NINA (Gain = 111, Offset = 10, Temp = -10°C). Also shoot new Bias Frames at -10°C using NINA (at Gain = 111, Offset = 10). Follow the preprocessing / Calibration steps outlined above. You should NOT need to modify the new Dark frames in any way at all. If they do not match amp glow patterns with your Light frames, then something else is still going wrong.

 

   After doing the above, I think the pattern may be much more subdued (or hopefully eliminated) in your integrated Light.

 

   Let us know how it goes.

 

 

John

Yes i am actually doing this right now. I just ran 35 dark frames at 4mins(240s) in NINA with the same parameters. I deleted those Darks from SHARPCAP.  So i can upload those to the drive account. Also running 60s darks as well. IDK how to check ADU #s either. Not a computerlectual lol



#24 Peregrinatum

Peregrinatum

    Surveyor 1

  • *****
  • Posts: 1,523
  • Joined: 27 Dec 2018
  • Loc: South Central Valley, Ca

Posted 01 October 2020 - 05:50 PM

How do the flats calibrate, do those curvy artifacts appear in the final image?  I have a mono camera that has funky flats like that and they calibrate really well.


Edited by Peregrinatum, 01 October 2020 - 05:50 PM.


#25 OrionNebula32

OrionNebula32

    Mariner 2

  • *****
  • topic starter
  • Posts: 235
  • Joined: 25 Dec 2017

Posted 01 October 2020 - 06:12 PM

Check my google drive, all the new darks are in there. 




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