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

flats not correcting dust spots

  • Please log in to reply
10 replies to this topic

#1 astromatto

astromatto

    Lift Off

  • -----
  • topic starter
  • Posts: 13
  • Joined: 09 Apr 2020

Posted 15 September 2021 - 10:35 AM

Hello everyone,
I am scratching my head since a week now and I still can't understand what is happening with my images, whatever lenght/ADU I take flats, it seems they have a dust spot but lights are not corrected cause I can't see the matching spot on the light frames.
I did 4 different sessions, taking 4 sets of flats, and all the flats show the same pattern => small gradient around the edges, small dust spot and big dust spot.

none of my uncalibrated light show the dust spots, after I calibrate those with the master flat tho, I see the dust spots confused1.gif

I started to wonder if, by any error, I shot without my filter, so today I took another series of flats, the big dust spot is gone, meaning it was on the filter, the small one is still there, meaning it was on the sensor; these flats leave me a bad gradient if used to correct the lights (flat_1700 in the folder)

I usually use PI to stack so I thought it maybe was one of the options so I tried to stack with Siril but I got the same identical result.

I also tried to lower the ADU I was trying to target for flats, same issue.

I run out of ideas, anyone had a similar issue and/or know how can I solve it? here the link to a folder where I put some of the frames https://drive.google...ZuaCWqFbXAaiBNk

I shot with a 533MC, astronomik cls-ccd and flats have been acquired with ekos automatically.


Cheers,
Mattia

Edited by astromatto, 15 September 2021 - 12:26 PM.


#2 pyrasanth

pyrasanth

    Aurora

  • *****
  • Posts: 4,597
  • Joined: 08 Jan 2016

Posted 15 September 2021 - 11:21 AM

I would take the following steps.

 

Remove the camera from the telescope. Cap the sensor and put the camera in a dark room and establish a connection to the computer and set the temperature. I would then take 500 bias subs- this is fast to do but takes a long time to integrate as a master- do nothing with these yet but put them all into a folder called bias. Put that folder into another folder called calibration.

 

Take another set of subs but this time your going to capture dark frames. The duration needs to be the same length as your light subs- so if your planning to shoot 1 minute light subs then you will need 1 minute darks. Take at least 30 darks and if possible 50 plus. Put these into a folder called darks in your calibration folder.

 

Clean your camera sensor and put it back on the telescope. Wait until evening when it is dark and take 60 flat frames at 50% of the max ADU of your camera. Put these into your calibration folder into a folder called flats.

 

Go to PI under scripts and open up the WBPP script. Click folder and point the script to your calibration folder. All the calibration frames will be added from that folder into the script. Configure the option to work with your CFA settings- i.e. RGGB settings or auto for the debayer type and anything under diagnostics that does not look okay. If you get stuck there is an excellent Adam block Youtube tutorial to explain the WBPP settings.

 

Run the script and it will correctly create your calibration files. The bias & dark master will be good for months- the light master as long as you don't move the camera.

 

What you have done is produced a set of bias, dak & flats which you will be able to use to calibrate your light subs. If you stick with WBPP and understand all its nuinsances you will be on your way to prodicing great images. I don't know what is wrong with your current calibration but starting from the beginning with these steps should fix the issue.

 

A lot of people think calibration is easy- it is not and requires care with the right number of calibration frames being taken to integrate and the correct ways to integrate the subs to form the masters. WBPP does this and removes the human error provided you put the right data types in.

 

One last thing- automation is great when you understand the process but can be bad if you don't understand what is happening. Do things manually- use the calibration flow charts in PI WBPP to see the calibration process visually. You will have a far greater understanding and a much better chance of understanding how to trouble shoot future issues.

 

Sometimes bringing the files from other applications into PI does not work as well as you would think. PI has everything you need to process an image so try as much as possible to stick with the one tool. It is okay to bring a nearly finished image out for post processing with say Photoshop but until then better to try & use PI as much as possible.


Edited by pyrasanth, 15 September 2021 - 11:32 AM.

  • BinoGuy likes this

#3 OldManSky

OldManSky

    Fly Me to the Moon

  • *****
  • Posts: 5,153
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 15 September 2021 - 11:51 AM

I downloaded one of your flats from what you shared above.

It's a little hard to tell, but I think it looks like the dust spot in the light is considerably smaller than in the flat.

Are they at the same focus point?  Did you change focus between the flats and the lights?

 

 

 

Attached Thumbnails

  • piflatissue.jpg

  • astromatto likes this

#4 astromatto

astromatto

    Lift Off

  • -----
  • topic starter
  • Posts: 13
  • Joined: 09 Apr 2020

Posted 15 September 2021 - 12:10 PM

You are right, the flat_1700 is the one I took without filter today. I used it to see if calibrating with it would solve my issue, but that wasn't the case, please ignore it, the other flats should match

#5 astromatto

astromatto

    Lift Off

  • -----
  • topic starter
  • Posts: 13
  • Joined: 09 Apr 2020

Posted 15 September 2021 - 02:19 PM

I would take the following steps.

 

Remove the camera from the telescope. Cap the sensor and put the camera in a dark room and establish a connection to the computer and set the temperature. I would then take 500 bias subs- this is fast to do but takes a long time to integrate as a master- do nothing with these yet but put them all into a folder called bias. Put that folder into another folder called calibration.

 

Take another set of subs but this time your going to capture dark frames. The duration needs to be the same length as your light subs- so if your planning to shoot 1 minute light subs then you will need 1 minute darks. Take at least 30 darks and if possible 50 plus. Put these into a folder called darks in your calibration folder.

 

Clean your camera sensor and put it back on the telescope. Wait until evening when it is dark and take 60 flat frames at 50% of the max ADU of your camera. Put these into your calibration folder into a folder called flats.

 

Go to PI under scripts and open up the WBPP script. Click folder and point the script to your calibration folder. All the calibration frames will be added from that folder into the script. Configure the option to work with your CFA settings- i.e. RGGB settings or auto for the debayer type and anything under diagnostics that does not look okay. If you get stuck there is an excellent Adam block Youtube tutorial to explain the WBPP settings.

 

Run the script and it will correctly create your calibration files. The bias & dark master will be good for months- the light master as long as you don't move the camera.

 

What you have done is produced a set of bias, dak & flats which you will be able to use to calibrate your light subs. If you stick with WBPP and understand all its nuinsances you will be on your way to prodicing great images. I don't know what is wrong with your current calibration but starting from the beginning with these steps should fix the issue.

 

A lot of people think calibration is easy- it is not and requires care with the right number of calibration frames being taken to integrate and the correct ways to integrate the subs to form the masters. WBPP does this and removes the human error provided you put the right data types in.

 

One last thing- automation is great when you understand the process but can be bad if you don't understand what is happening. Do things manually- use the calibration flow charts in PI WBPP to see the calibration process visually. You will have a far greater understanding and a much better chance of understanding how to trouble shoot future issues.

 

Sometimes bringing the files from other applications into PI does not work as well as you would think. PI has everything you need to process an image so try as much as possible to stick with the one tool. It is okay to bring a nearly finished image out for post processing with say Photoshop but until then better to try & use PI as much as possible.

hey thanks for the answer, couple of things, bias don't work very well with my camera and I am not taking them anymore, I use dark flats instead, bias were introducing noise and a master bias made from 150 frames looked like a flat pane, I fed WBPP with my lights, flats and dark flats, the result is a nigthmare



#6 pyrasanth

pyrasanth

    Aurora

  • *****
  • Posts: 4,597
  • Joined: 08 Jan 2016

Posted 15 September 2021 - 02:21 PM

"hey thanks for the answer, couple of things, bias don't work very well with my camera and I am not taking them anymore, I use dark flats instead, bias were introducing noise and a master bias made from 150 frames looked like a flat pane, I fed WBPP with my lights, flats and dark flats, the result is a nigthmare"

 

Then your doing something wrong.



#7 astromatto

astromatto

    Lift Off

  • -----
  • topic starter
  • Posts: 13
  • Joined: 09 Apr 2020

Posted 15 September 2021 - 03:23 PM

"hey thanks for the answer, couple of things, bias don't work very well with my camera and I am not taking them anymore, I use dark flats instead, bias were introducing noise and a master bias made from 150 frames looked like a flat pane, I fed WBPP with my lights, flats and dark flats, the result is a nigthmare"

 

Then your doing something wrong.

if for "you doing something wrong" you mean bias, I don't think is the case, I use dark flats for my reflex as well and it works very well. Dark flats contain the noise pattern usually found in the bias, that's why I am not using them.
If you are referring that I am doing something wrong with my flats now, that's possible, and that's why I am asking for help identifying my issue



#8 pyrasanth

pyrasanth

    Aurora

  • *****
  • Posts: 4,597
  • Joined: 08 Jan 2016

Posted 15 September 2021 - 04:30 PM

if for "you doing something wrong" you mean bias, I don't think is the case, I use dark flats for my reflex as well and it works very well. Dark flats contain the noise pattern usually found in the bias, that's why I am not using them.
If you are referring that I am doing something wrong with my flats now, that's possible, and that's why I am asking for help identifying my issue

If your saying that a bias won't calibrate your flats then something is very wrong indeed as a dark flat would probably have the same noise as a bias file. A dark flat with a very short exposure does not differ in many respects from a bias file.

 

By following a full calibration routine and learning to use the WBPP script correctly you will get to the route of the issue however its your choice if you follow the advice or not.

 

Good luck with resolving the issue.


Edited by pyrasanth, 15 September 2021 - 04:33 PM.


#9 astromatto

astromatto

    Lift Off

  • -----
  • topic starter
  • Posts: 13
  • Joined: 09 Apr 2020

Posted Yesterday, 03:07 AM

If your saying that a bias won't calibrate your flats then something is very wrong indeed as a dark flat would probably have the same noise as a bias file. A dark flat with a very short exposure does not differ in many respects from a bias file.

 

By following a full calibration routine and learning to use the WBPP script correctly you will get to the route of the issue however its your choice if you follow the advice or not.

 

Good luck with resolving the issue.

I think you misunderstood me a bit, I didn't say I am not gonna follow the advice, in fact I am preparing all the other calibration frames, I was just telling that from what I observed, bias frames for my sensor do not calibrate my images as there is no fixed pattern in them and I went for dark flats, also, I was never able to have PI scaling my darks to match my flats, that's an additional reason why I went woth dark flats



#10 sbharrat

sbharrat

    Viking 1

  • -----
  • Posts: 986
  • Joined: 28 Nov 2020
  • Loc: NJ, USA

Posted Yesterday, 04:13 PM

I think you misunderstood me a bit, I didn't say I am not gonna follow the advice, in fact I am preparing all the other calibration frames, I was just telling that from what I observed, bias frames for my sensor do not calibrate my images as there is no fixed pattern in them and I went for dark flats, also, I was never able to have PI scaling my darks to match my flats, that's an additional reason why I went woth dark flats

so dark flats definitely work more generally and do so even with cameras that have issues with very short exposure times (bias). However, the 533mc is not of those cameras that has an issue with bias (I have one of these). So it is odd that your bias didn't work if taken properly (which I think is what Pyrasanth was alluding to). 

 

As to your original problem, it is hard to know at this point. With the 533, it is reasonably straightforward to take flats but even then I still use a some sheets of paper to make my flats at about 1s long. My median ADU is about halfway. If you are already doing this, then either your imaging train changed somehow between the lights and flats, or the problem is in a different calibration frame. 

 

(Rather than trying to figure out if capture program is showing the original 14-bit number, or scaled to a 16-bit number, I took a l0 second exposure, looked at the maximum value in that, then aimed for an exposure that gave 1/2 that maximum.)


  • jdupton likes this

#11 jonnybravo0311

jonnybravo0311

    Vanguard

  • *****
  • Posts: 2,332
  • Joined: 05 Nov 2020
  • Loc: NJ, US

Posted Yesterday, 07:51 PM

Why would you be trying to scale your darks to match your flats?

 

With the 533, as Shaun stated, bias frames will work just fine to calibrate your flats. You can certainly use dark flats instead, but there really is no measurable difference. Sure, you might sacrifice fractions of an ADU here or there because of it. However, if you're having issues with taking biases and creating a master bias - which you stated earlier in the thread - you really should start your investigations there. Figure out what you're doing wrong in taking biases. In no situation, ever, should a master bias "look like a flat pane" as you indicated. A master bias should be nearly indistinguishable form a master dark when evaluated _especially_ if your dark flats are sub second exposures.

 

Instead of running through a script, try doing it manually. Go step by step and look at the results from each step. Isolate where things go wrong. Dropping a bunch of things into a script, flipping some switches and checking some boxes, while convenient, isn't doing much to help you understand where things are breaking.




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