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

ASI178mm camera and grid pattern seen in Ha

Astrophotography Solar CMOS Equipment Imaging
  • Please log in to reply
25 replies to this topic

#1 MER

MER

    Vostok 1

  • *****
  • topic starter
  • Posts: 197
  • Joined: 09 Nov 2020
  • Loc: Central, Ohio

Posted 25 April 2021 - 07:24 AM

So, as we all know that some ASI178mm cameras record a grid shaped pattern when used in Ha. There are numerous mentions on the net.

 

Also mentioned but rarely is the claim that the grid pattern can be eliminated buy using flat frames. I have gotten the feeling that some of these mentions were a theoretical response. So, I am looking for someone that has actually done this. 

 

I am looking for someone that has actually removed these grid patterns by using flats to do so. Someone that has personal experience with this issue.

 

 

 

Clear Skies,

Many Thanks!


  • SgrB2 likes this

#2 MalVeauX

MalVeauX

    Voyager 1

  • *****
  • Posts: 12,416
  • Joined: 25 Feb 2016
  • Loc: Florida

Posted 25 April 2021 - 07:39 AM

Hi,

 

It will be very challenging to eliminate the pattern with a flat, because it's in the flat too since its bayer matrix related. The grid can be eliminated manually with a routine that essentially corrects the bayer pattern so that it can be properly debayered. A user here demonstrated it and had results with it with personal code written in Linux. But he went silent a while back and hasn't responded to PM (user: jwestervelt).

 

Very best,


Edited by MalVeauX, 25 April 2021 - 08:17 AM.

  • SgrB2 and lpalbou like this

#3 EVS

EVS

    Vostok 1

  • *****
  • Posts: 131
  • Joined: 17 Jan 2009

Posted 25 April 2021 - 08:08 AM

I have been researching solar imaging for 3 months after replacing PST ITF filter as it rusted. The sun has become very active so I'm hoping the start imaging after I purchase a camera. Here is a link to CN topic on 178mm solar imaging you may find worth reading.
 

https://www.cloudyni...-alpha-imaging/



#4 SgrB2

SgrB2

    Soyuz

  • -----
  • Posts: 3,778
  • Joined: 10 Sep 2007
  • Loc: Ellicott City, MD

Posted 25 April 2021 - 10:41 AM

Prompted by Marty's remarks I stretched a flat that I took with the ASI178 mono

and see a grid pattern (see it below).  However, I did use this flat to produce the

animation images in this post:

 

https://www.cloudyni...ull-resolution/

 

I stretched one of the animation images that had the flat applied and could

not find any grid pattern. I might have not done justice to the stretch so if anyone

wants my data to satisfy their curiosity I am happy to provide them.

 

Cheers,

SgrB2

Attached Thumbnails

  • Annotation 2021-04-25 112525.jpg


#5 krakatoa1883

krakatoa1883

    Vostok 1

  • -----
  • Posts: 118
  • Joined: 02 Jun 2010
  • Loc: Mediolanum

Posted 25 April 2021 - 12:51 PM

So, as we all know that some ASI178mm cameras record a grid shaped pattern when used in Ha. There are numerous mentions on the net.

 

Also mentioned but rarely is the claim that the grid pattern can be eliminated buy using flat frames. I have gotten the feeling that some of these mentions were a theoretical response. So, I am looking for someone that has actually done this.

I usually remove the fixed pattern of my 178 via a flat frame through AS3! (taking a flat can actually improve images in other wavelengths too), it's quite easy to do and works fine for hires images of the chromosphere. However the pattern is not always present in the final image, it depends on filter, gain, sampling and other factors in a not-easy-to-predict way. For example I never seen it on images taken through my Lunt PT scope while it is often (but again not always) present when imaging through the Quark Chromosphere.

 

In post processing a slight denoise (I make use of this function in Astrosurface) applied before wavelets/UM/deconvolution can solve the problem in many (not all) cases.
 


  • SgrB2 and lorenzo italy like this

#6 lorenzo italy

lorenzo italy

    Viking 1

  • -----
  • Posts: 976
  • Joined: 04 Feb 2018
  • Loc: Montichiari (Bs) Italy

Posted 25 April 2021 - 02:50 PM

Just today this grid appeared to me.
I was shooting in white light, and I saw that if I used zero gain and minimum shutter speed, the grid would appear on the overexposed image of the sun disk.
I was using SharpCap, and the solar image was still overexposed because I was trying a filter for shooting in the cak line.
In addition to the Herschel prism I had only mounted the filter in question, but the light was still too much.
I then had to put an ND3, in order to have a useful shutter speed.
Next time I can, I take a picture of the grid.

 

Lorenzo


  • SgrB2 and lpalbou like this

#7 David Boulanger

David Boulanger

    Viking 1

  • *****
  • Posts: 845
  • Joined: 25 Aug 2020
  • Loc: Naples, Florida

Posted 25 April 2021 - 05:26 PM

I have a IMX 178 mono and can't say that I have seen this.  


  • Tapio likes this

#8 AJamesB

AJamesB

    Mariner 2

  • *****
  • Posts: 241
  • Joined: 17 Nov 2015
  • Loc: DFW, Texas

Posted 26 April 2021 - 02:10 AM

I've used the 178mm quite a bit, and fought this pattern noise quite a lot.  3 things helped me.

 

First, was using the "Pre-processing verticle and horizontal blur" option under 'Experimental features' in Autostakkert.  This adds a temporary blur to the images being aligned and stacked, essentially eliminating the finest details (you can adjust the amount) so that the fine bayer pattern gets blurred and then ignored during the stacking.  The blur is only temporary though and is not in the final stacked image.

 

2nd, was using 'drizzle' for the final stacked image.  This seemed to lessen the effect a little bit more yet. 

 

Even after those 2 steps, the bayer pattern was ultimately often the limiting factor in how much deconvolution I could apply to an image, especially images where the fine detail was on very small scales (like when using reducers), placing the fine detail on similar scales to the bayer matrix pattern noise. 

 

If, however, that wasn't the case, and the detail I was trying to pull out was on a larger scale than the bayer pattern, I could do a 3rd step, which was, in Pixinsight, zoom into the smallest levels of detail and using small settings in the deconvolution tool, do a fine scale bluring, both horizontally and veritcally, using a 1x5 pixel pattern.  This seemed to eliminate what remained of the pattern noise, albeit at the cost of the finest of detail, and allowed me to deconvolute the rest of the image a bit more than I could do without this third step.

 

Hope some of this helps.  I've since moved on to a different camera (asi290mm), and all though it doesn't have the same framerate as the 178mm, it also doesn't appear to have that pattern noise in spite of pixels of the same size (when in bin1), and I admit its really nice not having to worry about that pattern noise when I go to process images, or hesitating when reaching for that reducer as it forces the fine details down closer to the same scale as the bayer noise, lol.


Edited by AJamesB, 26 April 2021 - 03:32 AM.

  • SgrB2 likes this

#9 R Botero

R Botero

    Fly Me to the Moon

  • -----
  • Posts: 6,147
  • Joined: 02 Jan 2009
  • Loc: Kent, England

Posted 26 April 2021 - 03:50 AM

I was nervous about this when I was looking to upgrade my mono planetary camera (I have the ASI462MC for OSC) but went for the ASI178MM nevertheless as it matched the optics I use for Ha and CaK well.

 

I have had no issues yet with the grid when aligning/stacking or processing images.  I can see the pattern in low gamma, short exposure previews but it doesn't seem to be an issue for my stacking software (I use AS!3 but more often now AstroSurface, which is extremely fast).

 

My camera is relatively new - bought last month - so it could be that this issue depends on the batch of sensors ZWO have for production?

 

Roberto



#10 lorenzo italy

lorenzo italy

    Viking 1

  • -----
  • Posts: 976
  • Joined: 04 Feb 2018
  • Loc: Montichiari (Bs) Italy

Posted 26 April 2021 - 12:42 PM

 

 

My camera is relatively new - bought last month - so it could be that this issue depends on the batch of sensors ZWO have for production?

 

I do not know.
My 178mm is still three years old, and it shows the grille.

 

Lorenzo


Edited by lorenzo italy, 26 April 2021 - 12:43 PM.


#11 Aljo2345

Aljo2345

    Lift Off

  • -----
  • Posts: 20
  • Joined: 24 Jan 2021

Posted 29 April 2024 - 06:31 AM

Hi, there is a way to correct for the grid pattern. Before stacking, upload video file to PIPP, under the processing options tab check the "enable median noise filter" then process. The grid disappears form the video with no obvious degradation of video quality. Then stack as normal.


  • SgrB2 and MalVeauX like this

#12 SgrB2

SgrB2

    Soyuz

  • -----
  • Posts: 3,778
  • Joined: 10 Sep 2007
  • Loc: Ellicott City, MD

Posted 29 April 2024 - 10:27 AM

Hi, there is a way to correct for the grid pattern. Before stacking, upload video file to PIPP, under the processing options tab check the "enable median noise filter" then process. The grid disappears form the video with no obvious degradation of video quality. Then stack as normal.

 Is there a reliable download site?

 

 I found it:

 

https://astrowhat.co...cessor-pipp.38/


Edited by SgrB2, 29 April 2024 - 01:44 PM.


#13 SgrB2

SgrB2

    Soyuz

  • -----
  • Posts: 3,778
  • Joined: 10 Sep 2007
  • Loc: Ellicott City, MD

Posted 29 April 2024 - 12:10 PM

Hi, there is a way to correct for the grid pattern. Before stacking, upload video file to PIPP, under the processing options tab check the "enable median noise filter" then process. The grid disappears form the video with no obvious degradation of video quality. Then stack as normal.

 I tried it but it didn't work.  Here's the results which you must download

and zoom in to see the grid pattern in both the PIPP results with median

filter applied to the video then stacked and the Raw video then stacked.

The only way I know to kill the grid is to use a flat.

Attached Thumbnails

  • Screenshot 2024-04-29 125538.jpg

  • MalVeauX likes this

#14 gstrumol

gstrumol

    Fly Me to the Moon

  • *****
  • Moderators
  • Posts: 6,899
  • Joined: 03 Oct 2022
  • Loc: north of Detroit, Michigan USA

Posted 29 April 2024 - 12:54 PM

The grid is very apparent even in the small image, but certainly when you click on it. In both images.


  • SgrB2 and MalVeauX like this

#15 Aljo2345

Aljo2345

    Lift Off

  • -----
  • Posts: 20
  • Joined: 24 Jan 2021

Posted 29 April 2024 - 10:05 PM

 I tried it but it didn't work.  Here's the results which you must download

and zoom in to see the grid pattern in both the PIPP results with median

filter applied to the video then stacked and the Raw video then stacked.

The only way I know to kill the grid is to use a flat.

Hmm.. Interesting?

 

I will post shortly (I'm in Aus and at work atm) showing my results.

 

Are you processing in PIPP for 'surface detail'. I also converted file from SER to AVI?

 

What stacking software are you using?


Edited by Aljo2345, 29 April 2024 - 10:27 PM.


#16 SgrB2

SgrB2

    Soyuz

  • -----
  • Posts: 3,778
  • Joined: 10 Sep 2007
  • Loc: Ellicott City, MD

Posted 30 April 2024 - 03:12 AM

Hmm.. Interesting?

 

I will post shortly (I'm in Aus and at work atm) showing my results.

 

Are you processing in PIPP for 'surface detail'. I also converted file from SER to AVI?

 

What stacking software are you using?

 I used the surface detail option on the raw ser file and output was also

 a ser file.  To show the grid in both  I stretched both images with

 imppg using the exact same settings.  Stacking was done with

 AutoStakkert_3.0.14_x64.
 



#17 Aljo2345

Aljo2345

    Lift Off

  • -----
  • Posts: 20
  • Joined: 24 Jan 2021

Posted 30 April 2024 - 03:45 AM

 I used the surface detail option on the raw ser file and output was also

 a ser file.  To show the grid in both  I stretched both images with

 imppg using the exact same settings.  Stacking was done with

 AutoStakkert_3.0.14_x64.
 

Here are my 2 images, same wavelet settings for both, using Registax,

Attached Thumbnails

  • Sun 1 Jpeg.jpg

Edited by Aljo2345, 30 April 2024 - 03:53 AM.


#18 Aljo2345

Aljo2345

    Lift Off

  • -----
  • Posts: 20
  • Joined: 24 Jan 2021

Posted 30 April 2024 - 03:46 AM

Here are my images

Attached Thumbnails

  • Sun 2 Jpeg.jpg


#19 Aljo2345

Aljo2345

    Lift Off

  • -----
  • Posts: 20
  • Joined: 24 Jan 2021

Posted 02 May 2024 - 04:36 AM

Thanks for sending me your data SgrB2, to test.

 

The effectiveness of the PIPP's 'enable median noise filter' appears to be influenced by how much of the ASI178mm sensor is used: the smaller the area the less effective...  .

 

 

 

 

Attached Thumbnails

  • Sun3 Test.jpg


#20 Aljo2345

Aljo2345

    Lift Off

  • -----
  • Posts: 20
  • Joined: 24 Jan 2021

Posted 02 May 2024 - 04:43 AM

Here are a couple more of my images, using the whole sensor, before and after applying PIPP's 'enable median noise filter', then stacking.

Attached Thumbnails

  • Sun 1 and 2 small.png

  • MalVeauX likes this

#21 SgrB2

SgrB2

    Soyuz

  • -----
  • Posts: 3,778
  • Joined: 10 Sep 2007
  • Loc: Ellicott City, MD

Posted 02 May 2024 - 06:47 AM

The data below was stacked and processed with a flat

which was taken at nearly the same time as the data.

Such a flat will eliminate the grid pattern from the data

no matter the size or portion of the 178 sensor being

used for both the flat and the data.

Attached Thumbnails

  • 11_39_25_lapl7_ap889_conv.jpg

  • Aljo2345 likes this

#22 Aljo2345

Aljo2345

    Lift Off

  • -----
  • Posts: 20
  • Joined: 24 Jan 2021

Posted 04 May 2024 - 03:21 AM

Here are a couple more of my images, using the whole sensor, before and after applying PIPP's 'enable median noise filter', then stacking.

...and the result after some further processing.

Attached Thumbnails

  • Sun Small.jpg


#23 lpalbou

lpalbou

    Mariner 2

  • -----
  • Posts: 245
  • Joined: 22 Apr 2020
  • Loc: Paris, France

Posted 21 July 2024 - 02:29 PM

Prompted by Marty's remarks I stretched a flat that I took with the ASI178 mono

and see a grid pattern (see it below).  However, I did use this flat to produce the

animation images in this post:

 

https://www.cloudyni...ull-resolution/

 

I stretched one of the animation images that had the flat applied and could

not find any grid pattern. I might have not done justice to the stretch so if anyone

wants my data to satisfy their curiosity I am happy to provide them.

 

Cheers,

SgrB2

Was reading this as I just got the Player One Neptune M and I definitely have a grid pattern and even a line pattern issue. Your flat definitely shows them as well, the grid square is very visible..



#24 lpalbou

lpalbou

    Mariner 2

  • -----
  • Posts: 245
  • Joined: 22 Apr 2020
  • Loc: Paris, France

Posted 21 July 2024 - 03:10 PM

Just today this grid appeared to me.
I was shooting in white light, and I saw that if I used zero gain and minimum shutter speed, the grid would appear on the overexposed image of the sun disk.
I was using SharpCap, and the solar image was still overexposed because I was trying a filter for shooting in the cak line.
In addition to the Herschel prism I had only mounted the filter in question, but the light was still too much.
I then had to put an ND3, in order to have a useful shutter speed.
Next time I can, I take a picture of the grid.

 

Lorenzo

I do have that issue with a Lunt 40 and Player One Neptune M I just got (only visible when zooming) :
Lunt 40 + Player One Neptune M

 

However, I know gain can have a huge impact on sensor readout and pattern.. I was indeed at gain 0, so tomorrow (if sunny..), I will try with other settings


Edited by lpalbou, 21 July 2024 - 03:21 PM.


#25 lpalbou

lpalbou

    Mariner 2

  • -----
  • Posts: 245
  • Joined: 22 Apr 2020
  • Loc: Paris, France

Posted 21 July 2024 - 03:45 PM

AJamesB's solution to use AS4 experimental feature on blur horizontal/vertical in processing definitely helped to lower the grid/line effect:

 

178M grid / line pattern issue and AS4 fix
Left : normal AS4 + ImPPG
Right : enabling experimental feature / blur in pre-processing

Edited by lpalbou, 21 July 2024 - 03:45 PM.



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: Astrophotography, Solar, CMOS, Equipment, Imaging



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics