

NAME · USAGE · PARAMETERS · DESCRIPTION_ · EXAMPLES · TIME_REQUIREMENTS
BUGS · SEE_ALSO
nicreduce -- reduce a set of multiaccum images
nicreduce inputimages finalimage
- inputimages = ""
- Raw input images
- finalimage = ""
- Output combined image
- (darkfile = "niccal$skydark.fits")
- Dark frame (.fits). This can be a dark file taken
directly from the archive or one created by nicskydark.
- (flatfile = "niccal$flath.fits")
- A flat-field file, typically taken directly from the archive,
though some choose to make their own.
- (maskplfile = "niccal$badmask_2_mag.pl")
- Magnified bad-pixel mask frame (*.pl) used when combining images.
- (mask2file = "niccal$badmask_2.fits")
- Unmagnified badpixel mask used by nicfixpix (*.fits)
- (tempname = "temp")
- All the intermediate results will have this filename prefix
- (dotidy = no)
- Delete intermediate image files?
- (dodark = yes)
- Subtract dark frame? This step is mandatory.
- (doff= yes)
- Find pixel photon-rates using fullfitbam? This step is mandatory.
- (dobfix = yes)
- Correct bias error estimate (pedestal effect)? Mandatory if dofast=no.
- (donbias = yes)
- Correct bias levels in individual multiaccum readouts using nicbias
(pedestal effect). Mandatory if dofast=no.
- (doff= yes)
- Find pixel photon-rates using fullfitbam again? (pedestal effect).
This iteration uses the results of nicbias. Mandatory if dofast=no.
- (doflat = yes)
- Do flat-field correction? Mandatory.
- (domedian = yes)
- Do median filtering to remove striping? This optional step median
subtracts the median of each row from that row, and then likewise
for columns. This improves the flatness of the background. Not
recommended if you have large objects in your frames. Optional.
- (domagnif = yes)
- Magnify images times 2? Mandatory.
- (setmask = yes)
- Add spots to mask file? This step lets you interactively mask out
bad pixels. Must be done once.
- (domask = yes)
- Construct final mask? This step takes the list of mask spots and makes
the actual maskfiles. Mandatory.
- (setshift = yes)
- Set image shifts with imexam? Here you mark one object on each
of the frames. Must be done once.
- (doshift = yes)
- Make shifted data? This step shifts the images by integer pixel amounts.
Mandatory.
- (setxreg = yes)
- Set cross-correlation region and weight noise box? The final registration
of the images is done by cross-correlating a portion of the image.
Here you set that rectangular region, typically surrounding bright
objects in the image. Additionally you can specify a blank part of
the image to get a measurement of the noise. If you
specify a noise region, the images will be weighted based on the
noise. This is recommended if you are interested in detecting faint
objects, but not if you want the best bright point source images.
Must be done once.
- (doxreg = yes)
- Perform the cross correlation. Mandatory.
- (docomb = yes)
- Perform the final combination of the images. Mandatory.
- (interp = "spline3")
- Interpolation type used when magnifying images.
- (ff1sigma = 3.)
- Cosmic ray rejection threshold in sigma for fullfitbam1.
- (ff2sigma = 5.)
- Cosmic ray rejection threshold in sigma for fullfitbam2.
- (readnoise = 10.)
- Readnoise for fullfitbam. This is used for setting the cosmic
ray rejection thresholds.
- (z= 0., z2 = 2.)
- Brightness limits used when displaying images.
- (spotsize = 4.)
- Default spot radius in pixels when flagging bad pixels.
- (dofast = no)
- Skips the pedestal correction steps, i.e.
biasfix, nicbias, and the 2nd fullfitbam. If your data was taken before
21 August 1997, and you have large areas of blank sky in your frames,
you will want dofast=no. A flight software change on 21 August 97 reduced
the amplitude of the pedestal effect, so the pedestal correction
might not be necessary. If you have large extended objects in your
frames, the pedestal correction algorithm will fail, so you should
choose dofast=yes.
Nicreduce is the primary task of the nicred package.
It performs all of the steps necessary to take a set of raw, dithered
multiaccum images and turn them into a final combined image. Some nicred
users may choose to use nicreduce for reducing the individual images of
a dithered sequence and use their own software for combining the images.
Many of the parameters control whether a given step of the pipeline is
executed. Except where noted, all the steps must be executed to reach
the final combined image. If you want to repeat part of the reduction
process with a different calibration file (say you decide to change to
a different flat file), you can specify "no" for all the steps preceding
the flat correction. Certain steps (e.g. setting the bad pixel mask)
do not need to be repeated on subsequent reductions. These parameters
have names of the form "setXXXXX".
The parameters are listed in the same order in which the processing steps
are performed, so a look at the parameter list gives a good idea of what
happens. Following is a summary of the processing steps, the sub-tasks that
perform them, and the names of the files produced by each step.
Summary of the nicreduce pipeline:
Step Primary task Resulting file
------------- ------------ --------------
Subtract dark darksub _sub
Non-linearity correction
+fitting slopes+CR rejection fullfitbam _sub_s
Correct pedestal effect biasfix _sub_scorr
in _sub_s file
Use _sub_scorr to correct pedestal nicbias _bias
effect in each multiaccum
readout in _sub file.
Redo fullfitbam with pedestal fullfitbam _bias_s
effect corrected to improve
CR rejection.
Do flatfield correction nicflatten _flt
Interpolate over known bad pixels nicfixpix _fix
to avoid ringing in magnification
Magnify image magnify _mag
Shift each images by integer pixel amt imcombine _shift
to a common coordinate system.
This also fills in known bad pixels
with corresponding pixels from
other images.
Do fractional pixel shifting xregister _xreg
Combine images into final image imcombine final.fits
darksub, fullfitbam, biasfix, nicbias
Source Code · Search Form · STSDAS
Maintained by the Science Software Group at STScI
This file last updated on 27 May 1998