After some frustrating times trying to use PixInsight, I became disenchanted with the lack of documentation, the multiple ways to do what seem awfully much like the same thing, and I decided to give Siril another look. The tutorials looked interesting and they actually seem to explain concepts instead of offering a bunch of do this/do that recipes. But I come across a strange problem I don’t know what to with in background extraction.
The first reflex to have, and which allows you to properly see the gradient present in the image, is to switch to Histogram visualization mode. Thus, the image is visually very stretched and all the defects become visible.
Makes sense, but the examples given in the tutorial are monochrome.
Here is a stacked image I took of 66 5 minute images of the Whale Nebula, NGC4631, with my ZWO ASI533 MC Pro camera. NIght was moonless, light pollution very low. Stacking was with Siril script.
Well that sure looks odd. Where are all those crazy colors coming from? And how does this mode guide me in doing background extraction? All I can do is put in a bunch of samples just about everywhere except on stars, the main galaxy and that other galaxy below it.
The best I can get with Background Extraction is something like this (autostretch)
Not too bad but still more color than I’d like and it’s more vivid on my screen than it is here. and, when I tried to do color calibration it gave me an error message that my background extraction was wrong.
Hello, two things appear to be unusual with your images:
I’ve never seen an image turn blue after stacking, are you using the default settings for FITS debayering?
there is probably some stray lighting in your images, either calibration or raw, that makes these weird waves of colour. The background gradient removal tool may be able to get rid of that in the RBF mode, no need to place the points yourself, just use the default settings to see what happens
Background Gradient Removal tool. Is this the same as Background Extraction? I suppose so, there’s an RBF mode there and that’s what I’m using. The picture above shows the result of having done that (although I did edit the selected points).
So then, the question remains why the blue?
The individual frames to be stacked do not have that. I am using a version of the preprocessing script that has been edited to not use darks. I’d been told that darks aren’t necessary with my camera, a ZWO ASI533 MC Pro as that camera does not have amp glow. I know this topic is controversial, to put it mildly and I’m wondering if lack of darks could be the cause of the blue stacking result.
the default acquisition settings of ZWO cameras includes a white balance that doubles the blue, I’d say that’s what happens there. Open a raw image and look at the statistics in CFA mode for the 3 colors, if the median differs by more than 20%, it’s probably that
yes, what I called background gradient removal is background extraction.
So maybe a try at 50/50 would be worthwhile. This will have the effect of giving everything a green tint, which, of course, Siril has a process to remove. When I run that now, there isn’t much to remove.
Assuming this does what I want it to, that still leaves us with the odd histogram picture from my original post here
A global camera setting change seems unlikely to resolve this. What the heck might be causing that?
Hmm. I opened a flat in historgram mode (is that the same as “histogram equalization mode” and it comes up in black and white, not color. Is there a way to turn the color on?
