View on GitHub

IRAF Community Distribution

IRAF maintained by the community

Home | Installation | Packages | X11IRAF | PyRAF | Forum

bad pixel correction not being applied

Philip Massey wrote on Jan 09, 2009

Hi, Frank et al---

We are reducing some Magellan data, and after much head-scratching have concluded that neither ccdproc fixpix+ or fixpix is actually applying the bad
pixel mask. After nominally applying the bad pixel mask, the original and final image are identical, i.e., imstat reports a min and max value of 0 on the differenced image.
We are running 2.14.1, but I confirmed the same behavior on earlier
versions (i.e., 2.14 and 2.13-BETA2).

We have ccdtype set to "", and the other steps are proceeding fine; ccdlist shows the bad pixel mask as having been applied ("B") in the final image, and there are no errors reported. The bad pixel mask is listed in the header. But, nothing is actually happening, based on comparing the original and final images. This is consistent with the fact that we get a prompt return immediately.

In case this is something peculiar to our data, we've put the original image, the alledged fixed image, and the bad pixel mask where you can grab them.

Prior to fixpix:
http://www.lowell.edu/users/massey/original.fits
after ccdproc fixpix+
http://www.lowell.edu/users/massey/final.fits
bad pixel mask:
http://www.lowell.edu/users/massey/newbad.pl

thanks! We are really stuck until we can resolve this.
---phil

Philip Massey wrote on Jan 09, 2009

After some off-line correspondence with Frank, this problem was traced to the fact that the data I was trying to correct had not been taken with the same data region as the bad pixel mask. My goof!

Case closed...

Guest wrote on Jan 09, 2009

Hello,

Is the procedure for constructing and applying bad pixel mask exactly the same for echelle data as in the User’s Guide to CCD reductions? :? That is dividing the long and short exposure flats, using noao.imred.ccdred.ccdmask task and ccdproc with fixpix+.

Thanks in advance.

Siti

Philip Massey wrote on Jan 09, 2009

Bad pixels are intrinsic to the chip.
If there's a way to illuminate the chip without
the disperser (or without the cross-disperser) to make the bad pixel mask,
so much the better.

Guest wrote on Jan 09, 2009

Hello,

Thanks for observing tip. I need help with applying bad pixel correction to my echelle data using ccdproc task. There are two sets of long and short exposure blue flats. I did

cc> imarith lblueflat / sblueflat bflatdiv
cc> ccdmask bflatdiv badmap.pl
cc> imstat badmap.pl
#               IMAGE      NPIX      MEAN    STDDEV       MIN       MAX
               badmap.pl   2099200 0.1748     41.89        0.      28675.


I know this is not right. After displaying bflatdiv.fits, the bad pixels are in the rectangles defined by the following columns and rows:

30 30 1615 2050
180 180 1123 2050
64 64 1353 2050

There is the proto.text2mask task. Can someone show me how to do it? I find the help pages very confusing. My data file after trimming is of size [1024,2050].

Thanks
:) siti

Guest wrote on Jan 09, 2009

The file I constructed above can be applied as it is. I can see the bad columns disappear. :D

Last post on Jan 09, 2009