Combining different resolutions and FOV's

[chrisvdberge]

I know you can combine different FOV's/resolution images basically by scaling down the widest FOV to the narrow image. But can you also use it the other way around?
Reason I'm asking is I'll have the opportunity to image with 2 different refractors at the same time, while using a fullframe non-modified dslr (Nikon D600) and a modified crop sensor (D5100). So what I'm looking for is a way to basically shoot RGB with the D600 and H-alpha (using filter) with the D5100 and then combine the data.
Is there a way to do this so that you use the h-alpha data in the wide field RGB image? Or can you only go the other way?

[pfile]

you can - staralignment does not care about whether the reference is larger than the targets or vice versa.

one thing might be to try is to drizzle the smaller images so that you don't end up scaling up the non-drizzled subs too much. i have not tried this with DSLR images though - i think there is a script called BayerDrizzlePrep script that can help with this.

rob

[chrisvdberge]

Cool thx! I will try it out and see how it works out

[chrisvdberge]

So I'm currently on astrotrip in Namibia and trying out the combination of data of different scopes + camera's, but run into some problems;

The first night I had spacing problems with the flattener on one camera+scope, but still the images combined with just using the default settings in StarAlignment
However, on all the other data I shot after this the StarAlignment can't align the frames of one setup with the other. Funny thing is I fixed the spacing problem and therefor have a good flat fov which I thought would make it more easy to align the frames
One thing that's different as well is the camera rotation, in relation to each other (1 of them is rotated 180 degrees), but I can't imagine that this would be a problem?

I've tried different settings for the descriptors; triangles, pentagons, heptagons and upped the number of descriptors to 200. All to no effect.
It does find star pair matches (1428 using heptagons for the files linked below), but  the RANSAC  fails.
I've upped the RANSAC iterations to 8000 and tolerance to 4, all to no avail.
Anyone any idea what is going wrong here?

I've put 2 calibrated frames as example in dropbox:

Nikon D5100 with quadruplet:
https://www.dropbox.com/s/3hwaa6cu70ildxd/debayer_DSC_1827_c.xisf?dl=0

Nikon D600 with APM
https://www.dropbox.com/s/igfqoc80vhyukse/debayer_DSC_3213_c_r.xisf?dl=0

[pfile]

usually when this happens the star detection has latched onto hot pixels. there are a couple settings in SA to better reject hot pixels.

rob

[chrisvdberge]

not sure what I need to change for this to prevent. I've set the hot pixel removal to 1..

Here are the files as an example;

Nikon D5100 with quadruplet:
https://www.dropbox.com/s/3hwaa6cu70ildxd/debayer_DSC_1827_c.xisf?dl=0

Nikon D600 with APM
https://www.dropbox.com/s/igfqoc80vhyukse/debayer_DSC_3213_c_r.xisf?dl=0

[Juan Conejero]

Hi Chris,

Simply increase the noise reduction parameter to 1 in StarAlignment's Star Detection section. Your images are registered perfectly this way.

[Juan Conejero]

Your debayer_DSC_1827_c image has several problems. One of them is a lot of hot pixels, which have not been removed during calibration (you haven't used a master dark frame, right?). With debayering, hot pixels generate large artifacts that are very difficult to remove afterwards. Another problem is some kind of optical aberration on the left quarter of the image, which makes it very difficult to detect valid star centroids and also generates a strong distortion. Anyway, the images can be aligned with noise reduction.

[chrisvdberge]

Thx. May I ask what exact settings you use? Cause for me it is still not working..

re the hot pixels; this cam was modified and apparently there was some problem and the chip got damaged somewhat. With enough frames I get rid of them in stacking just fine, but apparently they are nasty in the registration