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

Strange pattern in flats

  • Please log in to reply
24 replies to this topic

#1 Shinken

Shinken

    Explorer 1

  • -----
  • topic starter
  • Posts: 72
  • Joined: 01 Jan 2020
  • Loc: Mallorca, Spain

Posted 23 February 2021 - 08:33 AM

I have noticed an unusual pattern in my flats. This does not happen in flats taken with R,G, B or Luminance filters but it happens with the Ha, Sii and Oiii filters. These are Astronomik 6 nm filters. It is more noticeable in the Ha filter and it messes up the calibration of the light frames as it will show in the light frame after the calibration is done. It looks like a pattern of many small squares measuring 16x16 pixels uniformly distributed throughout the image. I have seen other posts on CN showing the irregular blotches as you can see in the picture below taken with the Ha filter and showing the full frame. That larger scale pattern does not affect the calibration of the light frame but the reticular pattern does.

 

Ha1.jpg

 

Close ups.

 

Ha.jpg

 

It is a bit different for the Sii.

 

Sii.jpg

 

And the least noticeable for the Oiii.

 

Oiii.jpg

 

I am taking my flats using a light LED panel. It does not seem to matter if I use more or less power (brightness) in the panel. All flats were between 1 and 4 seconds as I recall. The orientation of the light panel has no effect. I do cover the light panel with a tight white t-shirt and I have tried to set a variable number of sheets of white paper between the panel and the scope but the pattern remains. That's why I feel that the pattern is internal somehow.

 

Has anyone experience similar problems?



#2 Shinken

Shinken

    Explorer 1

  • -----
  • topic starter
  • Posts: 72
  • Joined: 01 Jan 2020
  • Loc: Mallorca, Spain

Posted 23 February 2021 - 05:10 PM

Any suggestions?



#3 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 23 February 2021 - 05:16 PM

The only place I've really seen a pattern like this is with an OSC camera from the Bayer matrix.  But this is a mono camera, right?

 

The only other thing that comes to mind is a scaling artifact of some sort.  Are these right out of the camera, untouched?  Or have they be re-zed, binned, subsampled, or otherwise scaled during download?

 

-Dan



#4 clint.ivy

clint.ivy

    Sputnik

  • *****
  • Posts: 28
  • Joined: 30 Jul 2020
  • Loc: Chula Vista, CA

Posted 23 February 2021 - 05:21 PM

definitely looks like the CFA/Bayer pattern to me and we know that if you're using the WBPP process in PixInsight, it will integrate the master flat without debayering the flat subs.

 

So, back to Dan's question: Is your camera mono or One Shot Color?

 

**edit** it occurs to me that just today I debayered a master flat (trying to troubleshoot something else) and the grid pattern is still there! I wonder if there is some weird scaling happening when PI builds the master flat? Otherwise, it almost has to be coming from the LED panel - right??? (I'm using an LED tracing panel too)

 

defect In flats

Edited by clint.ivy, 23 February 2021 - 05:25 PM.


#5 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 23 February 2021 - 05:38 PM

definitely looks like the CFA/Bayer pattern to me and we know that if you're using the WBPP process in PixInsight, it will integrate the master flat without debayering the flat subs.

 

So, back to Dan's question: Is your camera mono or One Shot Color?

 

**edit** it occurs to me that just today I debayered a master flat (trying to troubleshoot something else) and the grid pattern is still there! I wonder if there is some weird scaling happening when PI builds the master flat? Otherwise, it almost has to be coming from the LED panel - right??? (I'm using an LED tracing panel too)

 

Don't see how it could come from the LED panel.  The pattern is very precise and seemingly in-focus.  The LED panel should be way out of focus.

 

-Dan


  • clint.ivy likes this

#6 milkychris

milkychris

    Vostok 1

  • *****
  • Posts: 102
  • Joined: 09 Jun 2019

Posted 23 February 2021 - 06:28 PM

I also use a led panel and see this pattern in my mono narrowband flats regardless of whether I use APP or PI to create the master flat.  I see it when using different programs to view the fits file.  It is always perfectly oriented with the frame so I assume it something microlens or sensor related and not from the led panel.



#7 clint.ivy

clint.ivy

    Sputnik

  • *****
  • Posts: 28
  • Joined: 30 Jul 2020
  • Loc: Chula Vista, CA

Posted 23 February 2021 - 06:31 PM

Interesting, I feel like this must negatively effect the final stack, and devalues the benefit of the flat. Thoughts? (Maybe I need to revisit the flats integration settings — I’ve ignored them and just let PI do its thing but maybe some large scale structure rejection is in order if available)

 

I also use a led panel and see this pattern in my mono narrowband flats regardless of whether I use APP or PI to create the master flat.  I see it when using different programs to view the fits file.  It is always perfectly oriented with the frame so I assume it something microlens or sensor related and not from the led panel.



#8 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 23 February 2021 - 07:10 PM

Can you share one or two of those images for a closer look?

 

-Dan



#9 clint.ivy

clint.ivy

    Sputnik

  • *****
  • Posts: 28
  • Joined: 30 Jul 2020
  • Loc: Chula Vista, CA

Posted 23 February 2021 - 11:00 PM

Dan, I believe your request is directed at the OP, but in case it’s not here’s my data https://www.dropbox....N3_VQrhZXa?dl=0 (see original post on why I’m looking at this, if curious https://www.cloudynights.com/topic/757494-is-this-defect-from-my-lens-dslr-both-or-other/)



#10 TxStars

TxStars

    Mercury-Atlas

  • *****
  • Posts: 2,570
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 23 February 2021 - 11:13 PM

A lot of LED panels have various patterns depending on the arrangement of the LED's and even how they are powered.

Most of the original applications for the LED panels do not require even illumination.



#11 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 23 February 2021 - 11:15 PM

Hi Clint:

 

Yeah I was asking the OP for a link to the full original images.

 

Your situation looks like a different problem.  Those streaks in the darks look pretty bad.  No idea what could be causing that.

 

-Dan



#12 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 23 February 2021 - 11:19 PM

A lot of LED panels have various patterns depending on the arrangement of the LED's and even how they are powered.

Most of the original applications for the LED panels do not require even illumination.

I think you're referring to LED studio lights used by photographers.  I have a few of those.  I've never seen anyone try to use those for a flat panel, but even so, it would be way out of focus and you would not see a pattern anything like that shown above.

 

LED tracing panels are normally used for flat fielding and they do produce a very even illumination.

 

-Dan



#13 TxStars

TxStars

    Mercury-Atlas

  • *****
  • Posts: 2,570
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 23 February 2021 - 11:42 PM

I'm talking about LED panels tracing or otherwise..

As I said many of them show patterns, some more than others.

@dan If you have one that shows no pattern please share which one you have.

 

To the OP  the pattern in the flats is in all the images you posted.


Edited by TxStars, 24 February 2021 - 12:11 AM.


#14 Shinken

Shinken

    Explorer 1

  • -----
  • topic starter
  • Posts: 72
  • Joined: 01 Jan 2020
  • Loc: Mallorca, Spain

Posted 24 February 2021 - 03:07 AM

I don't have a way to upload the large file at this time. I will look into it.

 

I am using a mono camera and processing the flat in PI as per the protocol every one uses. I do not think this is coming from the LED panel. As it has been said here, the pattern is very neat and not out of focus and as I said before I have rotated the panel in different position and it does not alter the pattern which should move around with moving the panel if that was the case. I think it is internal but I do not know where is it coming from or what to do to avoid it. It does present a huge problems as I may not be able to use flats now.



#15 TxStars

TxStars

    Mercury-Atlas

  • *****
  • Posts: 2,570
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 24 February 2021 - 05:28 AM

Have you tried using the flats as they are just to see how it comes out?

 

btw: what scope/lens camera combo is this?

 

A drop box of individual subs would help.


Edited by TxStars, 24 February 2021 - 05:35 AM.


#16 happylimpet

happylimpet

    Fly Me to the Moon

  • *****
  • Posts: 5,666
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 24 February 2021 - 05:43 AM

Some mono sensors have a bayer-type grid pattern as they are derived from colour sensors and the architecture reflects this. It all calibrates out fine.

 

In the case of my 1600MM, when I view at less than 100%, the sampling of the display software (ie not showing every pixel) makes this patterning visible as a sort of moire-interference pattern which consequently emerges ata  larger spatial frequency.

 

Upshot is, i see patterns like you show when i display my raw, flat and dark images at less than 100%. but it all calibrates out.

 

Are you sure you see these patterns in the calibrated image?

 

Also, when you view the flats (for example) at 100% scale do you see the pattern on the scale you describe, or just at the scale of the 2x2 'bayer' grid? If this is the case you can probably stop worrying.

 

I cant find my original post about this, but see for example:

 

https://www.cloudyni...id#entry8138738



#17 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 24 February 2021 - 09:40 AM

I'm talking about LED panels tracing or otherwise..

As I said many of them show patterns, some more than others.

@dan If you have one that shows no pattern please share which one you have.

 

To the OP  the pattern in the flats is in all the images you posted.

Here's the one I use:

https://www.amazon.c...e?ie=UTF8&psc=1

Although there are many different kinds and sizes on Amazon and none of them show a pattern.  That would defeat the purpose of a tracing panel, which is to have an even light source to trace over.

 

But again, the image would be so out of focus, it wouldn't matter if there was some kind of fine pattern.  Even T-shirts have a pattern created by the threads in the weave, but they are the gold standard for sky flats.  The thing you want to avoid for flats are large patterns or uneven brightness across the surface.  

 

The very fine pattern in the OP's images could not have been a pattern in the LED panel unless it was very far from the scope and the scope had been focused on it.

 

Then there is the fact that it is precisely aligned with edges of the image frame, and the rectangles are 16 pixels each.  That kind of precision and regularity would not occur if imaging a pattern on the LED panel.

 

-Dan



#18 Shinken

Shinken

    Explorer 1

  • -----
  • topic starter
  • Posts: 72
  • Joined: 01 Jan 2020
  • Loc: Mallorca, Spain

Posted 24 February 2021 - 10:33 AM

Here are the links to the original files:

 

Masterflat (Ha)

 

Individual flat image (Ha)

 

Calibrated light image (Ha)



#19 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 24 February 2021 - 12:20 PM

Thanks for sharing those.  Unfortunately, I'm not finding any more clues in them as to what's happening.  It looks like a sensor artifact to me, although I can't understand why it would show up in the flats and not the lights.  I would think it would be in both and be calibrated out.

 

-Dan



#20 Shinken

Shinken

    Explorer 1

  • -----
  • topic starter
  • Posts: 72
  • Joined: 01 Jan 2020
  • Loc: Mallorca, Spain

Posted 25 February 2021 - 04:07 AM

Thanks for looking at this issue. I don't know what else to do.



#21 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 25 February 2021 - 09:35 AM

I would contact ZWO and see what they say.

 

-Dan



#22 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,423
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 25 February 2021 - 09:38 AM

Since you say it only appears in your flats, It could be some sensor artifact that is related to exposure time.  So you might also try playing with various flat methods.  Try putting several layers of t-shirt or pillowcases over your flat panel so your exposures are longer.  Try taking sky flats.  Just try different light sources at different brightnesses to see if some work better than others.

 

-Dan



#23 clint.ivy

clint.ivy

    Sputnik

  • *****
  • Posts: 28
  • Joined: 30 Jul 2020
  • Loc: Chula Vista, CA

Posted 25 February 2021 - 09:52 AM

FWIW, I took a test set of flats yesterday with a different lens, 16mm, (troubleshooting another issue) and the pattern is still present so it almost has to be sensor artifact as Dan states. I would definitely try what Dan says (t-shirts).



#24 sharkmelley

sharkmelley

    Aurora

  • *****
  • Posts: 4,617
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 25 February 2021 - 10:23 AM

To the OP, you're not the first person to report the strange ASI183 grid-pattern phenomenon:

 

Mark


Edited by sharkmelley, 25 February 2021 - 10:55 AM.


#25 Shinken

Shinken

    Explorer 1

  • -----
  • topic starter
  • Posts: 72
  • Joined: 01 Jan 2020
  • Loc: Mallorca, Spain

Posted 25 February 2021 - 03:11 PM

Thank you, Mark. Yes. Those reports show the same pattern I am lloking at with my camera. It looks like is intrinsic to the sensor but should calibrate out. Would it be worth trying longer flats? I am around 2-3 seconds exposure.




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