Registering/aligning/scaling/cropping images WITHOUT stacking them

Greetings. We have a collaborative imaging project going on over at Cloudy Nights and I would like to know whether SiriL is capable of registering/aligning/scaling/cropping images WITHOUT stacking them to a master. We have a bunch of raw, unstretched stacks in FITS format from various scopes and cameras of a single target (the Rosette Nebula in this case), and I would like to register them to a chosen master stack (the reference stack) and produce a set of corresponding unstretched versions that all have same FOV, orientation, and image scale as the reference stack. Is this possible?

Failing that, is it possible to stack images of different FOV and resolution to a chosen reference with different weights? If so, how?

Thanks for your time.


Of course. Siril is not only a script runner. You can do approximately what you want in manual

Thank you; can you provide a workflow to accomplish this? Happy to RTFM, but any head start would be very helpful.

When I say manual I mean the opposite of automatic.
The link I gave you is the workflow.

I gave it a go and I don’t see any way to do what I asked about. I generated a sequence from all of the different sized frames, registered them, but see no way to export uniformly scaled versions of them. And when I tried to just stack them SiriL just closed.

You said ‘WITHOUT stacking them’ :smiley:
Stacking images of different sizes is not yet implemented in siril.
You can indeed calibrate the different sequences images and register them all, but for stacking, a crop will be needed. You could try seqcrop if you really want to do that with siril.
This is a feature that we expect to do in a future version.

1 Like

That’s correct. I want a set of images that are the same size and resolution as the chosen reference frame without them being stacked to a single frame. However, it seems that SiriL only produces those images, in passing, during the stacking process. Which is fine, as I can throw the stack itself away. I was able to get this process to work by excluding some of the input stacks from the squence, and one other failed to register and was not included. But at least 2 images of different sizes and resolutions were indeed registered and 2 aligned FITS files matching the reference frame in size and resolution were created, so in principle it can work. I’m not sure why it should work with input stacks and not others.

you are stacking stacks? You could process all the calibrated images together

I had the same question. I was looking for somthing like "Custom Rectangle in DSS. I also tried to make a sequence manually and after that stacking it. But no result. Perheps I miss somthing or do not understand it perfectly. IS there any help on this?

The new documentation contains what you need: Registration — Siril 1.2.0 documentation

Also all tutorials.

1 Like

Thanks a lot Lock042. I fanaly managed to understand a little more about how this works. And I succeeded in changing a sequence with a cropped reference frame. After stacking only the cropped area was stacked. Now I know that you first hae to set the working directory to the Process folder. That is where the seq files are living. The i thought I choose the r_pp__light_seq. I thought that was the most complete processed sequence. The I could bring up the Framelist. And view all the images to have a look for the best reference file, and choose that. In that particular image I selected my area of interest and cropped it. I was lokking if there was a way or a need to save those changes. I could not find that and just switched to the stacking tab and started the stacking. Works like a dream. My compliments to you and the Siril team. Now I go on with the rest of teh learning curve.