Page 1 of 1

debayering

PostPosted: Wed Mar 06, 2019 2:45 pm
by Rkonrad
As an alternative to debayering data in preparation for processing, would it be advantageous to load a dslr/one-shot tiff or fit as LRGB instead of just "open" (single file)? That is, loading it 4 times. It would seem to me that this would get rid of the double pixel green channel issue with the bayer matrix. I'm doing a presentation in a few days and though I want to keep it simple, I may want to bring this up if it improves colour accuracy in the end. Thanks!

Richard

Re: debayering

PostPosted: Mon Mar 11, 2019 11:13 pm
by admin
Rkonrad wrote:As an alternative to debayering data in preparation for processing, would it be advantageous to load a dslr/one-shot tiff or fit as LRGB instead of just "open" (single file)? That is, loading it 4 times. It would seem to me that this would get rid of the double pixel green channel issue with the bayer matrix. I'm doing a presentation in a few days and though I want to keep it simple, I may want to bring this up if it improves colour accuracy in the end. Thanks!

Richard


Hi Richard,

I'm not quite following. What's the "double pixel green channel issue"? Why would you want to avoid debayering? Are you talking a stack of images (which will have been debayered for stacking to work) or are you talking a single frame?
Loading a non-debayered dataset will just look like a mono dataset, and LRGB will treat it as such. Can you clarify? Thanks!

Re: debayering

PostPosted: Tue Mar 12, 2019 11:48 pm
by Rkonrad
Put more clearly (hopefully):

I have a stacked tiff from either a dslr or one-shot colour ccd. Instead of "opening" this file, I choose to load it under L, R, G, B thus loading it four times. Would this be similar to debayering the image? My understanding is that in the bayer matrix, there are 2 pixels assigned to green whereas only one for the other colours. I was thinking that the above process would equalize this.

Startools presentation will be better next time. It's really tough to slow down for beginners who know nothing about the software (or visual astronomers who are likely being polite :) ) I processed too good a data set. Next time I'll be more basic.

Cheers

Richard

Re: debayering

PostPosted: Fri Mar 15, 2019 2:47 am
by admin
Rkonrad wrote:Put more clearly (hopefully):

I have a stacked tiff from either a dslr or one-shot colour ccd. Instead of "opening" this file, I choose to load it under L, R, G, B thus loading it four times. Would this be similar to debayering the image? My understanding is that in the bayer matrix, there are 2 pixels assigned to green whereas only one for the other colours. I was thinking that the above process would equalize this.

The stacker debayers every individual frame, before aligning and stacking. Due to the changes in X and Y offset and rotation (amongst other things) between frames, it's impossible to stack still-bayered frames.
To get a better idea what a Bayer matrix is and how debayering works, have a look here. In short, StarTools doesn't really deal with Debayering, as that is the stacker's job.

Indeed, a Bayer matrix allocates 2 pixels out of 4 to the green channel. This makes the green channel 2x more precise after debayering (and is one of the reasons why StarTools asks what sort of image you're loading when using the Open button on the home screen). I'm not entirely sure what you mean by "equalizing" in this context though?

Startools presentation will be better next time. It's really tough to slow down for beginners who know nothing about the software (or visual astronomers who are likely being polite :) ) I processed too good a data set. Next time I'll be more basic.

You'll get there! :) I took me a couple of tries to figure out what sort of audience reacts best to what depth of information.