imalign artifacts

wfenton wrote on Jul 30, 2009

Please excuse if this is a silly issue. When I run imalign on a sequence of images the resulting shifted images show significant artifacts such as "holes" or dead pixels in the middle of stars. You can see a movie of before and after frames at the link below. Any thoughts on this?

http://files.me.com/billyfenton/9ls4e

---

Mike Fitzpatrick wrote on Jul 30, 2009

The link doesn't appear to be valid so I can't see what you're talking about. Assuming the shifts were all computed correctly this sounds like either some sort of interpolation artifact or possibly a data overflow. Are all the stars showing holes or only the really bright ones? You might try unlearning the task or change the interpolant type, you can also post the movie to ftp://iraf.noao.edu/pub so we can see what's going on.

-Mike

---

wfenton wrote on Jul 30, 2009

Thanks for looking. The two-frame movie is up in that directory and is called holes.mpg.

I notice that imal2 seems to always do integer shifts, is this best for photometry? I'm not sure how imalign works but am nervous about how it seems to be interpolating to get the shifts.

---

Last post on Jul 30, 2009