Comet stacking in Deep Sky Stacker. How does it do it?

[Ouroboros]

This has got me intrigued. I can see how standard stacking works in which the comet is fuzzy and the stars trail.

I can see how comet stacking works where the stars trail and the comet doesn't.

But I can't work out how the software achieves comet and star tracking - the "star freeze effect".

Anyone know?

[glappkaeft]

The general steps are:

* Align the image twice, one time aligned on the stars and one time aligned on the comet.

* Stack both images using methods that remove pixels that remove random pixls and moving objects (median, sigma clip etc.) .

* Of the two images you get from this one will remove the stars (fixed comet, moving stars)  and one will suppress the comet (fixed stars, moving comet).

* Align an merge the images.

[Ouroboros]

Thanks for attempting to explain how the software does it. I am not sure I'm much the wiser.

[damnut]

I think the best way is to try and stack with the advice above. It's easier to do than think about how it works :-)

[Ouroboros]

Yes, probably. I have managed to do use the star freeze effect as DSS call it. It didn't produce as nice a result as just stacking normally.

I'm just intrigued as to how the software does it. I have enough mathematical understanding to see how straight forward registering and stacking works, both on the stars or comet. I can't imagine the processing that enables these two images to be combined. A hand waving explanation is probably sufficient.

I once did a combination in Photoshop of a manually stacked comet picture against a manually stacked star field. It was possible to stitch the two images together because the region around the comet was conveniently free of stars. But I assume that's not how DSS do it.