Vignetting correcti...
 
Share:
Notifications
Clear all

Vignetting correction

Page 1 / 2

(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  

I'm trying APP using v1.081 trial. I watch every tutorial videos. In Mabula's vignetting correction video, he explains on an earlier version, draws boxes and just after clicking "calculate" button, APP corrects vignetting. That's very clear. But, on v1.081 a mask created after clicking calculate button, instead of correction. Seems like I have to do something with this mask, integrate on an image, merge an image or something else. What should I do with this mask?

vignetting screen

 

 


ReplyQuote
Topic Tags
(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  

In case I click on "show corrected image" following result appears:

vig

 


ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

Yes, the model is not right here, it’s very much over correcting. You did calibrate with bias and such?


ReplyQuote
(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  

@vincent-mod Thanks for reply. My calibration frames may be not perfect. I use QHYCCD163M first time, QHYFW3 first time, NINA first time and APP first time. There are many firsts. Flat frames were taken with white T-Shirt. Yesterday I just finished my DIY flat panel and took new flat frames. I hope they will help better.

But, With a vignetting correction tool, I was expecting an image, even if looks bad. Facing that kind of mask surprised me and would like to know if there is a bug in v.1.081.

My final image processed with APP without vignetting correction but remove light pollution tool is attached. As I said calibration frames especially flats are not ideal. I'll process same image with new taken flats.

Helix HSO1

 


ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

I think the vignetting tool is expecting a well corrected background as it uses that to determine a correct noise background and such (I’m not a mathematician though, so this is how far I dare to go 😉 ). It is a bit of a tricky tool though, I seem to get some over correction with it as well sometimes, it’s best to then use LP correction like you did here and indeed real proper flats.


ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

ps. great result for all those firsts!


ReplyQuote
(@tvlasits)
Hydrogen Atom Customer
Joined: 2 years ago
Posts: 2
 

Hi, I'm having the same problem, I would say it's definitely a bug, or some major weakness in the algorithm if it routinely over-compensates. Are the developers going to look into it?

 


SdA liked
ReplyQuote
(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  

I agree this is a bug. For sure Mabula will solve it.


ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

I don't think it's a bug, I've seen this been discussed before but can't find it unfortunately. I'll tag Mabula to see what he says about it (@mabula-admin).


ReplyQuote
(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  

@vincent-mod Dear Vincent, I do follow this topic to learn Mabula's answer and/or advices. If this is not a bug, what do I do wrong and how can I fix my mistake? On the other hand, in case this is a bug, is it working on it to correct? Because Mabula's tutorial on a previous version, vignetting correction works great. If other users have a clue and advice, I'm appreciated any help and advice. Thank you.


ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

This is an old post, but it also mentions it’s important to have a good correction and a workflow that might actually solve this issue; https://www.astropixelprocessor.com/community/tutorials-workflows/how-to/#post-449

Now, what might be wrong in the above posts I’m not sure (if it’s not the workflow), for that I would need to have a look at the calibration data. But I’ll also discuss this with Mabula, he’s a bit busy with the comet feature at the moment. @mabula-admin


ReplyQuote
(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  

@vincent-mod Thank you. I read all posts in the topic. I can sent flats, darks, biases and lights as sample if necessary. There are always a purple effect, which I can not fight only with light pollution tool, in my photos. I was hoping to correct it with vignetting correction but it's totally useless for me and really want to know why and fix my mistakes. Maybe I couldn't express my thoughts saying "purple effect". You can find 2 sample pictures below, perhaps they tell more clear than I did.

Lagoon Nebula HSO1 low

 

Helix nebula low res

ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

Ah ok, that’s a different question, might have more to do with the data. It’s best if I can take a closer look indeed.

If you can please upload some of the relevant subs to our server;

Go to https://upload.astropixelprocessor.com and use upload1 (or upload2 till 5) as username and password will then be the same as the username.

Create a directory named “volkanovun-pinkintegration” and upload in there. Thank you!


ReplyQuote
(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  

@vincent-mod Thank you. I uploaded darks, biases, dflats (Ha-SII-OIII), flats (Ha-SII-OIII) and lights 5 frames of each, under the directory volkanovun-pinkintegration. Please notice, lights completed in three nights, so each filter taken in different night. Because of that I checked "multi session" box while processing.

Thanks again


ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

Thank you! I'll download them tonight and hopefully have a clue tomorrow.


ReplyQuote
(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  

@vincent-mod Hi Vincent, could you take a look at the frames and have a clue?


ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

Of course, sorry it's busy down here I will get to it.


ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

Ok working on it, to start with.. I'm not seeing any extreme vignetting that would require the artificial flat. The regular flats seem to work fine.

I integrated the lights and used some LP correction on the Ha, SII and OIII and that looked fine. As this is narrowband data, the way the color comes out will be depending on how you combine the data in the RGBCombine tool. Narrowband data alone won't have "natural" colors as it's not broadband RGB data, so you would need to pick a combination that you like and tweak the colors manually and will depend on your liking. I'm guessing you combined the above in the tool with the RGB formula?

I did have to crop it a lot as the frames were not the same.

Screenshot from 2020 08 05 11 05 14

ReplyQuote
(@vincent-mod)
Quasar Admin
Joined: 5 years ago
Posts: 4723
 

For instance here I used a different combination formula (R: 50% H-alpha, 50% SII, G: 20% H-alpha, 80% OIII, B: 100% OIII).

Screenshot from 2020 08 05 11 13 32

ReplyQuote
(@volkanovun)
Main Sequence Star Customer
Joined: 2 years ago
Posts: 43
Topic starter  
Posted by: @vincent-mod Ok working on it, to start with.. I'm not seeing any extreme vignetting that would require the artificial flat. The regular flats seem to work fine.

 I'm guessing you combined the above in the tool with the RGB formula?

Thanks for your effort. No, not RGB I combined HSO1 formula. I was expecting a little more and distinct, clear blue arond the core. A purple color become instead blue. On the other hand, I don't know how to set percentage H-a, OIII, SII...


ReplyQuote
Page 1 / 2
Share: