Combine RGB Images/...
 
Share:
Notifications
Clear all

19 June 2021: Our upload server https://upload.astropixelprocessor.com/ has been migrated successfully to our new office with higher upload and download speeds (nearly 10MByte/sec up/down ) ! We now have 1 general upload user called: upload with password: upload. The users upload1 - upload5 have been disabled.

31 May 2021: APP 1.083-beta2 has been released ! APP 1.083 stable will follow soon afterwards. It includes a completely new Star Reducer Tool, New File Saver Module, Improved Comet registration and much more, check the release notes here!

DOWNLOADS are available HERE!

 

Combine RGB Images/Frame size  

Page 2 / 2
  RSS

(@wvreeven)
Quasar Admin
Joined: 3 years ago
Posts: 1122
August 5, 2021 15:14  

@dalepenkala Hi Dale,

The bayer or CFA pattern normally is embedded in the FITS files so you indeed should not change anything in tab 0. APP is capable of detecting the CFA pattern by itself.

Wouter


ReplyQuote
(@dalepenkala)
Main Sequence Star Customer
Joined: 4 months ago
Posts: 31
August 5, 2021 18:27  
Posted by: @wvreeven

@dalepenkala Hi Dale,

The bayer or CFA pattern normally is embedded in the FITS files so you indeed should not change anything in tab 0. APP is capable of detecting the CFA pattern by itself.

Wouter

Ok thank you Wouter! I now have a computer problem that is stopping me from going further now so I’ll let you know how things work out once I get my computer back up and running!

 

Dale


ReplyQuote
(@wvreeven)
Quasar Admin
Joined: 3 years ago
Posts: 1122
August 19, 2021 11:15  

@dalepenkala Dale and I have been discussing his data offline and I have created a workflow for him that I'd like to share here.

I followed the following workflow to extract al the color channels from your data and combine them again. Note that I needed to use APP 1.083-beta1 because of an issue with the star registration in 1.083-beta2.

This is the workflow I followed to get the IDAS Ha and OIII images:
1) In tab 1 load the 240 IDAS lights, the IDAS Flats, the 240 Darks, the Dark Flats and the Bias
2) In tab 0 for pattern select RGGB, enable force Bayer/X-trans CFA and for algorithm select Ha-OIII extract Ha
3) In tab 6 click integrate and provide the name for the object: Veil-Ha
4) In tab 0 for algorithm select Ha-OIII extract OIII
5) In tab 6 click integrate again and this time provide the name: Veil-OIII

Next restart APP because that is easier than clearing out all loaded frames. So, now we will do the L-Pro data as follows:
1) In tab 1 load the 180 L-Pro lights, the L-Pro flats, the 180 Darks, the Dark Flats and the Bias
2) In tab 0 for pattern select RGGB and enable force Bayer/X-trans CFA. You can leave the algorithm as is since the best one for RGB automatically gets chosen when APP is started.
3) In tab 6 click integrate and provide the name: Veil-RGB

Now that we have the Ha, OIII and RGB images, we need to do two things. One thing is to align them to each other and the other is to extract the R, G and B channel from the RGB image. The order in which these two things are done doesn't really matter. Once again, close and reopen APP. Then load the Ha, OII and RGB images as lights in tab 1. APP will state that these look like processed images and will ask if you are sure. Yes you are. Next, go to tab 4 and click "start registration". Once it is done, click "save registered frames". Make sure to save them as 32 bit float FITS files. Now the frames will be aligned to each other and will have the same dimensions. Close and restart APP again. Now load the registered RGB image as light. Go to tab 2, enable "split channels" and click "SAVE (calibrated) Light frames". Again, make sure to save them as 32 bit float FITS files.

Now your data is ready to be combined. For the last time, close and restart APP. Go to the tools tab and load the registered Ha, OIII, R, G and B files. Note that APP will not recognise what channels the Ha and OIII files are so you need to select those by hand. Make sure to select the RGBHOO formula and click the "new formula" button. Then click "(re-)calculate" and enjoy the result.

Screenshot 2021 08 19 at 11.13.57

ReplyQuote
(@wvreeven)
Quasar Admin
Joined: 3 years ago
Posts: 1122
August 19, 2021 11:17  

@dalepenkala I should add that this result can be improved. This can mostly be done by applying light pollution correction on the Ha, OIII and RGB files before registering them. Also normalizing them to each other in tab 5 right after registering the three files will help. I'll leave it to Dale to try this himself 🙂


ReplyQuote
(@dalepenkala)
Main Sequence Star Customer
Joined: 4 months ago
Posts: 31
August 19, 2021 15:46  

@wvreeven

Hello Wouter,

Thank you so much for this! I will give all this a try based on this workflow! I do have a question though.

Your statement on tab “0” In tab 0 for pattern select RRGB

I’ve been using RGGB for my OSC camera as thats what I was always told the pixel structure is. Is this really what I should be using? So I should be using the RRGB and NOT the RGGB.

I’m going to mention something that Wouter and I talked about offline for EVERYONE to read/know and that is I didn’t realize that I should be using the FITS file structure for ALL my data and calibration files. I have been using the Uncompressed file TIFF. I always was told that this was the best file format to use because it was uncompressed data. This just shows you stupid I am with all this computer/tech stuff! In the end I’ll be reshooting ALL my calibration library in the coming days!

Again, I’d like to thank Wouter, for his patience here and offline as well as all of this information! It has taught me a lot!

I’m sure I’ll still have questions Wouter so I hope that I can still bug you once and a while 🙄

Now I have to put this all into practice!

Thanks!

Dale


ReplyQuote
(@dalepenkala)
Main Sequence Star Customer
Joined: 4 months ago
Posts: 31
August 19, 2021 15:51  

Wouter, you mentioned you had to use the “beta1” because there is an issue with the star calibration part of beta2. I have noticed that when I use the star calibration the results are drastically different and in fact commented on this in other forums without any answers. Is this what your talking about in beta2 is the drastic color change to the image.

Thanks!

Dale


ReplyQuote
(@wvreeven)
Quasar Admin
Joined: 3 years ago
Posts: 1122
August 19, 2021 18:39  
Posted by: @dalepenkala

Your statement on tab “0” In tab 0 for pattern select RRGB

I’ve been using RGGB for my OSC camera as thats what I was always told the pixel structure is. Is this really what I should be using? So I should be using the RRGB and NOT the RGGB.

Thanks for spotting that typo. I corrected it. It indeed should be RGGB.


ReplyQuote
(@wvreeven)
Quasar Admin
Joined: 3 years ago
Posts: 1122
August 19, 2021 18:42  
Posted by: @dalepenkala

Wouter, you mentioned you had to use the “beta1” because there is an issue with the star calibration part of beta2. I have noticed that when I use the star calibration the results are drastically different and in fact commented on this in other forums without any answers. Is this what your talking about in beta2 is the drastic color change to the image.

 

Read my post again. I wrote that there is an issue with the star analysis. This is step 3 of the integration workflow of APP and is done automatically when going from tab 1 to tab 6 and clicking the integrate button. For me the star analysis and then the registration (tab 4) of the frames failed for all IDAS frames but these steps go OK when using 1.083-beta1.

I don't know about the color change that you are referring to. I posted my results of my workflow above and the colors look natural to me.


ReplyQuote
(@dalepenkala)
Main Sequence Star Customer
Joined: 4 months ago
Posts: 31
August 19, 2021 20:16  

@wvreeven

Ahhh ok Wouter my bad, miss read your comment there, sorry!

Yes I agree the colors look very normal to me as well. I think I'm going to download the 1.083-beta1 just to have it in case I run into problems as you stated.

 

Thanks!

Dale


ReplyQuote
Page 2 / 2
Share: