How to correctly debayer raw data

[lensman57]

Hi,

I have a question with regards to debayering, in PIPP there are a few options for debayering. How can one tell if the software has choosen the correct one as there are a few options for debayering manually too. How can we be sure that we have debayered correctly and does this really make a difference?

Regards,

A.G

[IanL]

Yes it makes a big difference. The easiest way to tell if you have the right setting is to take a daylight picture with the same camera (if you can) and then load/debayer that. Just try the different options until you get the right colours.

Even if you can't do that you can usually tell if you have the right settting by trying them all on an astro-image. If you use the wrong setting you will usually get an odd colour-cast, too blue, red, green or similar. For most cameras you can also look up online what the bayer pattern is and set it manually.

[lensman57]

Yes it makes a big difference. The easiest way to tell if you have the right setting is to take a daylight picture with the same camera (if you can) and then load/debayer that. Just try the different options until you get the right colours.

Even if you can't do that you can usually tell if you have the right settting by trying them all on an astro-image. If you use the wrong setting you will usually get an odd colour-cast, too blue, red, green or similar. For most cameras you can also look up online what the bayer pattern is and set it manually.

Thank you , will give it a try.

Regards

A.G

[cgarry]

With PIPP you only need to worry about getting the debayer pattern right, just leave the debayer algorithm at the default 'Bilinear'. There are only 4 debayer patterns to choose from and as IanL suggested the easiest way to find the correct one for your camera is to try them all on a daylight image. Though it it normally obvious on astro images too.

Cheers,

Chris

[billhinge]

I found this description yesterday http://www.mathworks.co.uk/help/images/ref/demosaic.html

I opened up a RAW image in Pixinsight and set the DSLR RAW formatter not to do a debayer (Pixinsight uses the opensource dcraw library so you could do this stand alone with dcraw or other program using dcraw if you didn't have Pixinsight). Having got my undebayered image I zoomed in till I could see the bayer matrix and compared it with the web site.

The other way mentioned above also works, I found if you picked a couple of the totally wrong options you don't get the bayer pattern removed correctly, or another wrong option has the red and blue the wrong way around

[lensman57]

I found this description yesterday http://www.mathworks...f/demosaic.html

I opened up a RAW image in Pixinsight and set the DSLR RAW formatter not to do a debayer (Pixinsight uses the opensource dcraw library so you could do this stand alone with dcraw or other program using dcraw if you didn't have Pixinsight). Having got my undebayered image I zoomed in till I could see the bayer matrix and compared it with the web site.

The other way mentioned above also works, I found if you picked a couple of the totally wrong options you don't get the bayer pattern removed correctly, or another wrong option has the red and blue the wrong way around

Hi,

Thanks to all of you for your advice, my confusion arises from the point that lets say if I alter the debayering pattern in PiPP, there doesn't appear to make any difference or any that I could see, there is also some options in RegiStax under pre filter during alignment with the same result.

Regards,

A.G

[cgarry]

Are you debayering an image or a video in PIPP?

A few thoughts:

* PIPP will not debayer if it detects that the frame is already in colour.

* You need to press the 'Test Options' button after changing any options to update the image preview.

* The debayer pattern cannot be changed for raw images as that is automatically selected by the DCRAW code.

Cheers,

Chris