======================================================================
SN9621 MMT reduction -- R.Knop

Taken files 19mar00xx where xx=34..41 from
/home/astro22/deepsearch/mmt_mar96

Presuming that these files have already been "ccdproc"'ed and flatfielded,
based on header comments.

Image files: 36 and 39

Remove cosmic rays: use pct_CR_rem.e ... ... -100 3000 F F 3 2 250 T

Do it iteratively: first 36 with 39 as ref, then 39 with 36 as ref.
Repeat 2 times... still lots of junk left over.

Try writing my own version -- badpix5.  Runs on Orchid files.  Does a
similar thing with a source and a reference, only the way I've written it it
seems to be really excited about finding bad pixels.  To avoid getting rid
of every slight noise spike, badpix5 gets rid of pixels whose difference
from the corresponding pixel in the reference image is greater than a
siglimit times the standard deviation of differences in a nxn box.  There is
a single siglimit; there is no distinction between pixels in the open and
next to an already identified cosmic ray.

This program probably needs some work, but for now run it with a siglimit of
7sigma.  This seems to work fairly well, although it's still zapping a touch
more than it ought to.  It's not out of hand though.  (With siglimit=4, it's
out of hand.)

>> cr0036, cr0039 ; image with bad pixels in mask0036, mask0039

...doesn't look very flatfielded.  There is a bow in the sky along the
spatial axis!

Now, as to identifying objects -- there seem to be a number.  ROTANG=-3.9 in
the header; if that is really a slit position angle of +3.9 degrees, then
the objects which show up on the slit are consistent with images.  The SN
pops in right at about pixel 195-200.  There are two at higher pixel values,
adn three more objects down between pixels 140 and 170.  There are objects
in a line at ~4 degrees with the same relative spacing on the max light
image.  (I can't check the absolute spacing because I haven't figured out
the scale of either the image or the spatial axis of the spectrum.)

Rough IRAF pixel positions (by eye): 148,155,167,199*,204,209

Identify the wavelengths:
 cr0036.000?: order=2 spline3 to 11 lines in 19mar0035 ; resid RMS=0.303pix
 cr0039.000?: order=2 spline3 to 13 lines in 19mar0040 ; resid RMS=0.251pix
    delete the 7503 line, which seems to muck things up

Flux/atmosphere/etc calibration -- use "Feige 34", frame 19mar0023 (already
ccdproc'ed), which in turn is wavelength calibrated with 19mar0024

  >>dc0023.0001  (Feige 34)

Call the "flux file" flux0023 and the sensfunc sens0023.0001