PDS_VERSION_ID        = PDS3
RECORD_TYPE           = STREAM
OBJECT                = TEXT
  PUBLICATION_DATE    = 2001-11-01
  NOTE                = "Anomalies and discrepancies in the CD Volume set"
END_OBJECT            = TEXT
END

   COMMENTS ON THE STARDUST NAVIGATION CAMERA EARLY CRUISE IMAGES

This document contains comments concerning anomolies regarding this
archive:

  1)  At the time of the release of this archive, the geometric
  calibration of the camera was not known with absolute confidence.
  Because of this fact, the TWIST keyword is "UNK" (for unknown) for
  each of the images on the archive.

  2) One of the images is incomplete because the data was not
  completely transmitted to the ground, namely N0110C101.IMG.  The
  portion of this image which was transmitted shows a completely over
  exposed frame.  The decision was made to discontinue the downlink of
  the remainder of the data for that image to save bandwidth.
  
  3)  The FRAME_SEQUENCE numbers are not strictly monotonically
  increasing.  Two gaps exist in the numbering scheme.  The known gaps
  are from and including frame 0004 up to and including 0099, as well
  as frame 0109.  The decision was made at the beginning of the
  project to start experiments at a 'round' number.  This decision was
  reversed at the time of the planning for frame 0110.   After these
  experiments, all frames have been consecutive.

  4)  The image labels in the Early Cruise Images dataset contain keywords
  that have not been approved for use by the PDS.  Future dataset
  volumes will either have these keywords removed or approved.
  
  5)  The IMAGE_HISTOGRAM object in the Early Cruise Images dataset is
  missing the required ITEM_BYTES keyword.  It has ITEM_BITS instead.
  
  6)  Some of the images have WINDOW subobjects in their labels.  This
  object type was in negotiation with the PDS at the time they were
  created.  The keyword WINDOW_DESC should be DESCRIPTION.
  
  7)  In the Early Cruise Images labels, the constant 1.0E32 was used
  instead of the  string "UNK", and -1.0E32 was used instead of the
  string "N/A".  Future datasets will use the strings instead.
  
  8)  In the Early Cruise Images, some images labels specify a TARGET
  based on the intentions of a test, and not a true target.  There
  are no target catalog files for these targets.  The TARGET keyword
  will not be used in this manner for future datasets.

  9)  In image N0110C101.IMG, the IMAGE CHECKSUM was a negative number.
  This image was fully saturated, and overflowed a software counter.
  This has been corrected an will not happen for future data.