Problem with square/block artifacts processing Comet A3

Don't use SXT. Also, don't drizzle beyond 1x.

Thanks for the reply, but how can you use comet align on images with stars, when you do that the stars streak out as the comet becomes aligned. As I understand it you can only comet align starless images? Please can you expand on your suggestion?
Also, the comet aligned subs are 'registered' individual subs, they are not the drizzled master light? I don't understand this point, I wonder if you can explain a bit further maybe?

Tim BUCKLEY:
Thanks for the reply, but how can you use comet align on images with stars, when you do that the stars streak out as the comet becomes aligned. As I understand it you can only comet align starless images? Please can you expand on your suggestion?
Also, the comet aligned subs are 'registered' individual subs, they are not the drizzled master light? I don't understand this point, I wonder if you can explain a bit further maybe?

You can use comet alignment also with stars. Actually, I myselt just did today. However, as you mentioned, you get steaked stars. I tried and could remove the streaked stars with SXT (and also had no blocks by applying the large overlap). Still, there were a few streaked stars left over, but I could remove them by hand.

Klaus Haaken:

Tim BUCKLEY:
Thanks for the reply, but how can you use comet align on images with stars, when you do that the stars streak out as the comet becomes aligned. As I understand it you can only comet align starless images? Please can you expand on your suggestion?
Also, the comet aligned subs are 'registered' individual subs, they are not the drizzled master light? I don't understand this point, I wonder if you can explain a bit further maybe?

You can use comet alignment also with stars. Actually, I myselt just did today. However, as you mentioned, you get steaked stars. I tried and could remove the streaked stars with SXT (and also had no blocks by applying the large overlap). Still, there were a few streaked stars left over, but I could remove them by hand.

*** Thanks, I will try this ***

What's the problem with streaked stars? It's the way all long exposure comet images used to look. I know it's not pretty but it's an accurate portrayal of the object.

Jonathan Hankey:
SXT does appear to be the problem, I have the same issue and found other's online reporting the same thing.

Starnet++ works but it's much more tedious since it has no bulk operations.

Also how do you get just the stars out? Create a star mask and invert?

Create the stacked-up comet-registered image with a strong rejection filter to obtain a comet-only image (you might have to remove by hand the stars leftover,s if needs be). Then use that in CometAlignment in the "Operand" argument. It is all explained (for once!) in the contextual help. You'll get the stacked up star only image. Add the two together.

I've had this when drizzling once. Used the non drizzle file and all was perfect.

Make sure you are using the most efficient StarX model, that it is updated, and that you have large overlap enabled. Make sure negative values aren’t being created when calibrating the images (adequate pedestals), as the comparatively shorter exposures many take for comets can lead to relatively dark exposures.

The tiling comes from the regions in the image StarX is sampling and can *usually* be fully mitigated with those ducks in a row.

If that doesn’t do the trick, you will probably want to use an older comet workflow which relies on outlier rejection in stacking rather than star removal from individual subs. This tends to leave a little more artifacting in stars, but might just need to be what is done. Or if Starnet works and StarX won’t cooperate, that may be the way to go. Just as long as you can take measures to protect the core of the comet.

Just curious…once the frames are properly aligned, have you tried removing the stars from each of the aligned images, then integrate? I actually don't have a lot of experience here, but was playing around with ways to align some frames I acquired recently on Tsuchinshan-ATLAS, and this was one of the ideas I had. I suspect its probably tedious, but, I also suspect that star removal would work better?

Chris Ashford:

Jon Rista:
Just curious...once the frames are properly aligned, have you tried removing the stars from each of the aligned images, then integrate? I actually don't have a lot of experience here, but was playing around with ways to align some frames I acquired recently on Tsuchinshan-ATLAS, and this was one of the ideas I had. I suspect it’s probably tedious, but, I also suspect that star removal would work better?

Jon, yes, that’s what I am doing and I assume others in this post thread are doing the same. I use StarXterminator in batch mode with a mask to protect the comet nucleus on all my CometAligned subs. Once that’s done ImageIntegration to stack the starless comet subs.

CometAlignment can also be used to subtract the comet from the registered subs by using the integrated starless comet as an operand.

Stack the stars, then blend the comet back in, and everything starts to come together nicely…..

Ok...interesting...

Comet A3
Following the advice/link from @Tony Jerig I integrated the comet aligned subs but WITH STARS, I then cranked up the high rejection on the Integration drop downs - rejection 2. I did apply a pass of SXT large overlap on the integrated image but not convinced this was entirely necessary.