Drizzle and dual ba...
 
Share:
Notifications
Clear all

[Solved] Drizzle and dual band filters

Page 2 / 2

(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 

Dear Robert, @rbarbout

Thank you very much ! We have received the upload in good order 😉

I will work on this issue in the coming days and will let you know my findings.

Mabula

This post was modified 3 months ago by Mabula-Admin

ReplyQuote
(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 

Hi Robert @rbarbout,

My first testing indicates that there is NO BUG here with the O3 extract and Bayer Drizzle combination. The artefacts that you get are a combination of 2 things it seems:

- too little data for droplet size of 1.0. If I set droplets to 2.0, all artefacts are gone and the result is similar to interpolation instead of Bayer Drizzle. This is what we would expect.

- the blue channel is clearly worse than the green channel in terms of signal. So in your setup with your camera and filter more O3 signal is caught in the green channel compared to the blue channel.

These 2 factors combined result in the artefacts that you see. The drizzle droplets of green and blue do not match and can't average properly due to too little data with droplets of 1.0.

And this does explain why the single Green And Blue Bayer Drizzle channels look okay when extracted from the color Bayer Drizzle result.

I will run a couple of more tests. Maybe a solution for this issue will be to do a normalization of blue and green before making the O3 mono signal from Blue and Green Combined. In that way, even with too little data, the too small droplets can still average nicely together and not give drizzle artefacts. To be clear, what you see in this issue is a typical drizzle artefact resulting from too little data for the set drizzle parameters. (So yes, i think the solution if any would be in how green and blue are combined to get the mono O3 result)

Mabula

 

This post was modified 3 months ago 2 times by Mabula-Admin

ReplyQuote
(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 

Hi Robert @rbarbout,

Good news, i have done further testing and I did find 2 bugs now 🙂 

  1. the data normalization was not working correctly in case of Bayer Drizzle for monochrome Narrowband OSC data.
  2. for O3 and Ha-O3 extract O3 data, the green and blue channels were not combined properly.

I am fixing both issues now so it will work properly starting with 2.0.0-beta5. I will share results on your data as soon as both issues are fixed.

Mabula

 


ReplyQuote
(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 

Hi @ennui2342, @bjoernh, @rbarbout

I have fixed the Bayer Drizzle HaO3 extract O3 (or just O3) processing. Robert's data helped me to find and solve 2 bugs now.

  1. the data normalization was not working correctly in case of Bayer Drizzle for monochrome Narrowband OSC data.
  2. for O3 and Ha-O3 extract O3 data, the green and blue channels were not combined properly.

Both are now fixed. The artefacts in the O3 are not there anymore and the overal result is much better:

The Ha and O3 extract results from Robert's data after implementing the fixes:

HaO3 O3 BayerDrizzle integration
HaO3 Ha BayerDrizzle integration

Zoomed in to see the pixels and possible artefacts:

HaO3 O3 BayerDrizzle integration zoom in
HaO3 Ha BayerDrizzle integration zoom in

And a preview of the Ha O3 bi-color that can be made from this:

HaO3 composite from BayerDrizzle integrations

Thank you all for reporting this and thank you Robert for uploading good test data so I could fix this properly now.

Mabula


ReplyQuote
(@rbarbout)
White Dwarf Customer
Joined: 3 years ago
Posts: 24
 

@mabula-admin 

Very good news indeed !

I was still puzzled since I made another attempt with a stack of 91 subs and the artifacts were still present.

I am now reassured and wait impatiently for the next beta.

Thank you for addressing this problem so quickly.

Robert.


ReplyQuote
(@ennui2342)
White Dwarf Customer
Joined: 4 years ago
Posts: 18
Topic starter  

Fantastic work Mabula, I'll dig out the old data from this thread and give it a run through for comparison!

-Mark


ReplyQuote
(@dav78)
Main Sequence Star Customer
Joined: 4 years ago
Posts: 83
 

Yes, it's a good news indeed! Thanx @mabula-admin.

I've noticed this issue since at least one year, but haven't reported it as I though it was something relative to my datas, and I didn't use this function anyway.

I certainly will now with 2.0 👍 

 

David


ReplyQuote
(@rbarbout)
White Dwarf Customer
Joined: 3 years ago
Posts: 24
 

Hello @mabula-admin 

So I conducted some tests with the pre-beta5 and unfortunately, the checkerboard pattern is still present when extracting OIII in Bayer-Drizzle mode.
I did not see any difference between result from beta4 except the beta5 pixel values are now twice the value of beta4.
I used the very same data I already sent you and took some screnshots to show you the results:

Zoom

Top center is Green channel from HaOIII_color in bayer drizzle mode
Top right is Blue channel from HaOIII_color in bayer drizzle mode
As you can see, no checkerboard pattern in these images even if they come from a bayer drizzle intergration of just 30 subframes.
To build a OIII image, I mixed the Green and Blue channels with formula (2G+B)/3. I called this image GGB.
As you can see in top left image, again, no pattern visible.

Bottom left is the result from HaOIII_extract_OIII in bayer drizzle mode . Here the checkerboard pattern appears.
As experiment, I have subtracted this OIII image from my synthetic GGB image. The results are in center and right bottom images (GGB-OIII and OIII-GGB).
These images show the differences between the extracted OIII and a OIII created by mixing Green and Blue channels. The lighter the pixel, the more the difference.
The checkerboard appears clearly indicating evenly spaced deviation in pixels values.
These deviation cannot be caused by lack of subframes since the Green and Blue channels do not have these artifacts and moreover, I conducted the same tests with another image of 95 subs and the result is the same.

I hope these data will help you to find what's going on. During this time, I will keep my GGB image as OIII even if it is not as accurate as the extracted OIII.
Below are the same images with a lower magnification.

Capture

Last point, Red channel from HaOIII_color and Ha image from HaOIII_extract_Ha are the same and none is showing the checkerbord pattern so it seems the problem is just when combining the blue a nd green channels.
All settings to default except stars detection set to 1000 and background neutralization disabled
All images taken with an ASI294MC Pro (CFA pattern = RGGB) integrated in Bayer Drizzle mode, scale 1, droplet size 1, topHat Kernel.

Robert


ReplyQuote
(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 

Hi Robert @rbarbout,

Thank you very much for your detailed response. I re-opened this issue so I will work on this first thing tomorrow to see if I can solve it now... one question though for now: why the GGB image, what argument do you have for such a synthetic over a GB synthetic ?

Mabula

This post was modified 2 months ago by Mabula-Admin

ReplyQuote
(@rbarbout)
White Dwarf Customer
Joined: 3 years ago
Posts: 24
 

@mabula-admin 

Good question !

I had no idea how to build an OIII from green and blue images. Then, I found it was the way an other application builds its OIII.

Maybe it is linked with the fact the Bayer matrix has two green pixel for one blue ?

Or to the fact the green response of an ASI294MC with a L-extreme filter is twice the blue response…

Or I am completly wrong and I should have mixed 50% of each !

Robert

This post was modified 2 months ago 2 times by Robert Barboutie

ReplyQuote
(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 

Hi Robert @rbarbout,

Let me first start by thanking you for pointing out this issue so clearly. I should have really fixed it now 🙂 and I think I fully understand why the artefacts occured.

The O3 channel reconstruction in 1.083 and 2.0.0-beta4 was indeed exactly the cause of the issue. Our reconstruction method was too simple.

We need to realize that the response of the O3 signal in the green and blue channels will always be different per optical setup due to :

  1. the actual filter used (can be any Duo/Triple/Quad NB or an O3 filter which all can have different filter characteristics)
  2. the actual camera's sensor QE sensitivity on these wavelengths...
  3. the filter characteristics of the Bayer filters (is expected to be different for diffent camera models and brands)

Because of these 3 reasons, I would never argue that it would be logical to create an O3 reconstructions using ( 2*G + B )/ 3 = O3.

It think it is more exact to try to find the actual scale factor between the response in the blue channel versus the green channel for the incoming O3 signal. This is what I implemented now and I can no longer see any checkerboard pattern like before. Noise drops clearly when compared to the 2.0.0-beta4 (beta5) version and thus SNR increases significantly. It really looks much better now!

This new fix is also relevant to the HaO3 mono result where Ha and O3 are combined to 1 signal.

Results now with the 2.0.0-beta6 version when compared to 2.0.0-beta4 (which is identical to 1.083 stable for this issue) for your data:


O3 extract results zoomed in to pixel peep, left is beta6, right is beta4:

200beta6 O3 extract NO artefacts
200beta4 O3 extract artefacts

beta6 Signal to Noise Ratio (SNR) : 3,55

beta4 Signal to Noise Ratio (SNR) : 2,66

So SNR is improved siginificantly!


HaO3 mono results zoomed in on same region, again left beta6 and right beta4:

200beta6 HaO3 mono NO artefacts
200beta4 HaO3 mono artefacts

beta6 Signal to Noise Ratio (SNR) : 4.34

beta4 Signal to Noise Ratio (SNR) : 3.02

Again, SNR is improved siginificantly!


Or in this 4 panel comparison:

BayerDrizzle Narrowband comparison

On which platform (windows, linux, macOS intel or arm) do you use APP Robert? I will provide a beta6 installer with this fix so you can hopefully confirm that you see the same improvements 😉

Mabula


Dav78 liked
ReplyQuote
(@rbarbout)
White Dwarf Customer
Joined: 3 years ago
Posts: 24
 

Hi Mabula @mabula-admin ,

Congratulation, I think you nailed it !
From the images you provided it seems you solved the problem.
I will be happy to check the fix with a "pre beta" !
I use APP on a PC running Windows 10 64bits.

Robert


ReplyQuote
(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 

Hi Robert @rbarbout,

🙂 In the meantime, we decided that we will go ahead and release 2.0.0-beta6 for all users today or tomorrow with this fix. I will send a link with the windows installer to your email adress once I have created it though so you can test quickly if you have time.

Mabula


ReplyQuote
(@rbarbout)
White Dwarf Customer
Joined: 3 years ago
Posts: 24
 

OK @mabula-admin ,

Should be able to give it a try this evening (Paris time).


ReplyQuote
(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 
Posted by: @rbarbout

OK @mabula-admin ,

Should be able to give it a try this evening (Paris time).

Hi Robert @rbarbout,

This is APP 2.0.0-beta6 for windows with the fix 😉 please let us know if all is okay now !

Mabula

 


ReplyQuote
(@rbarbout)
White Dwarf Customer
Joined: 3 years ago
Posts: 24
 

@mabula-admin ,

Perfect ! No more problem with OIII in bayer drizzle integration.
Tested with scale = 1.0 or 2.0, droplet size = 1.0 or 1.5
It's a GO for beta 6 !

Thank you Mabula and the APP team.


ReplyQuote
(@mabula-admin)
Quasar Admin
Joined: 6 years ago
Posts: 3321
 

Excellent Robert @rbarbout 😊,

Thank you very much for testing so quickly !

Mabula


ReplyQuote
Page 2 / 2
Share: