Description of the calibration corrections to fix the error introduced        
by differences in the shutter opening direction.                              
                                                                              
                                                                              
Problem                                                                       
------------                                                                  

The double-bladed shutter utilized by the camera has a delay in its
slide mechanism that introduces an offset in the actual exposure time,
compared to the commanded exposure time.  It was originally reported
that exposures with the forward shutter direction were 1.65 ms shorter
than the commanded exposure time, while images with the reverse
shutter direction were as commanded.  During the recalibration of the
NAVCAM instrument as part of the Stardust NExT extended mission, it
was found that the original offsets were incorrect.  During the
recalibration (which will be fully documented in the Stardust NExT
archive) it was found that offsets are present in both the forward
and reverse directions.  For exposures with a forward shutter
direction, the exposure time is 0.40 ms longer than the commanded
time, while those in the reverse direction have an exposure time 
1.34 ms shorter than the commanded time.

It was also previously reported that the parity (where all even frame
numbers have one shutter direction and odd numbers have the other)
only changed when a bias frame was taken, since the shutter blades
didn't fire.  It was subsequently found, however, that the shutter
resets to the forward direction when the NAVCAM power is cycled,
potentially changing the parity in a sequence of images.  Furthermore,
there are occasions when an image wasn't obtained for some reason,
which also has the potential for changing the parity.  Descriptions of
the images are included in the DATASET.CAT file and the PIIMGLOG.ASC
document, where information can be found about bias frames and
potentially missed images. 

                                                                              
Questions about the Parity
-----------------------------                             

A list of all Stardust images is included in the file
SHUTTER_PARITY_LIST.ASC in the documents directory, interspersed with
the times of NAVCAM power cycles and notes about bias frames and
potentially missed images.  The parity of each image, as well as can
be determined, is noted in this document.  There are still
ambiguities, where it is not known if an image was skipped (thus
changing the parity) or if it was obtained but not downloaded (thus
maintaining the parity).  The information in this document is used to
determine how to apply the exposure corrections with regards to the
shutter direction.

As can be seen in this document, the shutter direction is known for
most of the images.  However, there are several groups of data where a
"missing" image introduces an ambiguity in the shutter direction of
subsequent images (e.g., it is not known whether the shutter triggered
or not).  Fortunately, most of the ambiguous sequences consist of long
exposure images (>1000 ms).  Exposures this long are essentially
unaffected by the difference in actual and commanded exposure time
(the <0.1% effect is well within the calibration errors of the
instrument).

The one potential exception to this case involves the Anne Frank
images, where only every third image was saved in the frames 374
through 426.  It is possible that the shutter did not fire on one of
the unsaved images, changing the parity.  Tests performed at the SBN
suggest that this is not the case, and that the shutter direction
listed in the SHUTTER_PARITY_LIST.ASC is indeed correct, though this
is not absolutely confirmed.  (It should be noted that these images
are not calibrated, and many of them are partially saturated,
minimizing the problems related to the true exposure time.)


Exposure Time Corrections                              
--------------------------

Because of the ambiguous long-exposure images, combined with the fact
that the shutter exposure correction is negligible for long exposures,
it was decided that, for consistency, corrections would only be
applied to the shorter exposures where they actually affect the
photometric measurements.  Thus, the exposure times for all images
with commanded exposure times shorter than 1000 ms have been corrected
for the shutter timing offsets.  

The following corrections are applied to each image as relevant, given
the information in the SHUTTER_PARITY_LIST.ASC document.  For the
following steps, the commanded exposure time is converted to the
actual exposure time by an increase of 0.4 ms for images with the 'F'
designation, and a decrease of 1.34 ms for images denoted 'R'.
                                                                             
1) The EXPOSURE keyword in the FITS header is corrected to the actual 
   exposure time.      

2) The EXP_CMND keyword is added to the FITS header to denote the
   originally commanded exposure time.

3) The EXPOSURE_DURATION keyword in the PDS label is corrected to the actual 
   exposure time.
                          
4) In the datasets containing calibrated images, the flux levels are
   adjusted to account for the difference in exposure times (e.g. flux
   levels are multiplied by t_cmnd/t_actual).