Aligning RGB frames to Luminance frames ?

[Tycho65]

Hi SkySearchers,
I have started mono imaging with a mono camera but have not had much chance to practice due to poor weather since about Sept 2023. I have got some limited access to a remote scope and can only get limited numbers of exposures due to access/cost reasons.

Please can someone explain if this is normal. I have 1 frame each of L,R,G,B, all binned 2x2. The L was imaged from a different night, but the RGB the same session/night. In fact I think the L from the same night was of very poor quality (I can ask for advice another time). I got a screen shot of this when I did the RGB composition in Siril of the (calibrated RGB frames).

The R,G, and B frames are slightly misaligned. The camera is mono and the telescope is a 17 inch CDK. Is this normal ?
Normally, I might stack and align the R, G, B separately to a reference frame in the Luminance filter for example. But I only have 1 frame of each. I have actually managed to use ASTAP to align the frames but ASTAP changes the black and white points after the alignment which I do not like. I also find the ASTAP hard to use, due to the large number of options and things you have to choose and I have to work out which is the correct ones so I do not spend time trying various options etc.

Here is a screenshot of the RGB image when trying to compose in Siril (autoalign does not work in Siril).
Question: Are RGB frames normally mis aligned like this ? (? due to refraction) or is this a problem with the remote observatory?
Many thanks everyone.

Magnus

[Graeme1858]

I've seen something similar to this once before (I'm quite new to LRGB imaging myself!) The cause was, after stacking, I combined the LRGB frames prior to stretching them. I started again and stretched the individual LRGB images, then did an LRGBCombination and the problem was corrected.

Graeme

[Tycho65]

Hi Graeme,

So do you have to process each R,G,B frame (like background extraction, neutralisation, noise reduction,and histogram stretch - how much stretcing?) before (L)RGB combinatiion? When do you do Deconvolution?

Thanks.
Magnus

[Graeme1858]

So do you have to process each R,G,B frame (like background extraction, neutralisation, noise reduction,and histogram stretch - how much stretcing?) before (L)RGB combinatiion? When do you do Deconvolution?

Yes, for each of the four colours up to the stage of a Masked Stretch or EZ Soft Stretch (if you haven't updated!) (Linear Stage) (I used to use Screen Transfer Function/Histogram Transformation) Then combine the four images. Then do the Curves, Saturation, Star Reduction etc (Non Linear).

Everybody has their own way so I'm sure others will have different workflows. Can't remember where Deconvolution fits in. I used Cuiv the lazy Geek's video:





We will be reposting the AP Processing Challenge some time soon, so you can join us there and get some practice.

Graeme

[KathyNS]

If the problem is just that the different frames are misaligned, that's not a problem: it is normal. You will almost never have all frames come out of the camera perfectly aligned. The registration step of stacking is supposed to correct that. If registration is not aligning them, then it must be seeing some hot pixels that look more star-like than the actual stars.

If the problem is that you can't stack them because you only have one frame of each, there are two ways around that (besides taking more frames, which is the preferred solution). I have had automatic stacking procedures complain that I didn't have enough frames to stack. It is a warning that is usually not to be ignored, but if you want to press ahead and do it anyway, just duplicate the single frame. If it wants at least three frames of each filter, make two duplicates. It won't know that they are all the same; it just counts them.

The other way around having too few frames to stack is to do all the stacking steps by hand. The math for the calibration step is a pain, but do-able. Registration will register whatever you tell it to.

[Tycho65]

Thanks Kathy,
I managed to get Siril to register and align the files, but the final RGB showed slightly misaligned stars, perhaps because the stars in each R,G,B were of different sizes, perhaps slight misalignment.
I think I will need more files as you said to get decent results.

Magnus