Colour channel combination - cant seem to get right

[Srgilbert69]

Hi all - new to PixInsight and have been working through the examples and tutorials around the place and trying to workout a workflow. I hope I am just missing a piece of the puzzle somewhere. I have been imaging in narrowband (HA, O3, SI) and  I go through all the preprocessing steps and it all seems to be coming along nicely and I end up with 3 frames that are all registered against each other and linear fitted to each other. I convert to non linear and then try to combine but can never get anything that looks remotely good.

Attached is a screen shot from the other night, the left hand images are the integrated images, the middle column is these images after dynamic background extraction and stretched to non-linear and the image by itself is the channels combined.


Any help here would be greatly appreciated.

Thanks

Shane

[pfile]

something's wrong here - the input images don't seem to be different narrowband wavelengths. are you sure your filter wheel was not stuck on one filter or something?

rob

[Srgilbert69]

Thanks for the reply. I diid check the filter wheel was rotating and filters were in the right way, so I hope I got that right. Will check it all again though as normally something is failing on me

Attached are 3 frames straight from the camera, just autostretched. Subtle difference in the 3. New to this so still not sure what looks right and what doesn't. 

Thanks again for the reply.

Cheers

Shane

[ngc1535]

A linear fit is performed between images?
This will tend to give them equal weight (in signal and neutral background).
It just seems that weighting the channels as required wouldn't include a linear fit.
In the case of broadband weight determination, PCC is the standard. I am not certain what the method of choice is for NB color mapping-
but LF between the channels seems like you are making the contribution from each channel
more like one another... and therefore less color variation?

This is a good topic to learn from! Can't "weight" for the answer!

-adam

[pfile]

ok those 3 look like they came from different filters.

i usually do linear fit the SII and OII to the Ha... but then the following processing probably undoes that. i usually just follow what rick outlined here when he introduced ColorMask:

https://pixinsight.com/forum/index.php?topic=7751.0

i think the easiest thing would be if you posted the 3 masters; it is very difficult to make conclusions from screenshots...

rob

[Srgilbert69]

Thanks again for the reply -

I have been following the tutorial

https://www.lightvortexastronomy.com/tutorial-narrowband-hubble-palette.html

I have started again from scratch with a fresh set of darks taken last night and am trying it on a different set of lights that have a bit more contrast in them so will see how I  go there.

I will post the 3 at the end of the pre processing stage.

Thanks

Shane

[Srgilbert69]

Real newbie question now - with the attachment size limited to 512KB  how do I upload the masters ?

Cheers

Shane

[pfile]

you'll need to upload them to something like google drive, or onedrive or dropbox, make the folder shared, and post the link here.

rob

[Srgilbert69]

Thanks Rob - thought that might be the case but thought I would check. Am uploading the pre processed images to a drop box but I did everything again from scratch and got a much better result this time. So it seems like I must have made the same mistake with both sets of data I tried with. 

Attached it the eagle nebula which previously didn't combine either for me. Will go back and try the other one again and see if I get better results.

Seems to be all user error so I probably wont post the drop box link to the image data unless you are interested.

Cheers

Shane

[pfile]

that looks right for an SHO narrowband image. glad it is solved.

rob

[John_Gill]

Hi,

Instead of doing a "ChannelCombination", I use LRGBCombination because you set individual Channels Weights, adjust Lightness & Saturation and Chrominance Noise Reduction.

space is not black
John