The below e-mail exchanges offer an explanation of the JJT database that was delivered in four parts on 9_15_04.  The complete JJT Database with ALL of the JJT metadata was delivered after we had reached a project stopping point to scan habs/haer/hals photographs.  The four parts of the database are named:
habshaer1.xls - 117 MB
habshaer2.xls - 121 MB
habshaer3.xls - 114 MB
habshaer4.xls - 117 MB

hhJJTdbFinal9_2004Summry1234Combnd.mdb - 29 MB
This is a more accessible database that combines key fields ("summary") from all of the above four databases.  

- Kit Peterson

>>> Katherine A. Peterson 9/22/2004 6:28:30 PM >>>
Phil,

After additional explorations, (and exporting and merging a 10 key field abbreviated version of the four Excel tables into a single Access table so that I could more easily query the entire thing), I believe that the database as delivered is fine and that records for all of the images that we have received are contained in the database.  The database also contains records for all of the 216 batches that were delivered.  

Some statistics:
HH Final & Complete JJT Database: 259,163 records
HH Final JJT image delivery Database: 231,946 (27,221 fewer records)

HH Final & Complete JJT Database - Records with a Batch # associated: 232,006 (60 images less then image delivery database = 2 CDs)  I could explore this further through analysis and comparison of the CDVolumes...

The ADDITIONAL 27,221 records in the JJT database need to be analyzed and accounted for.  Some of these include records for surveys which do not even have photos.  Some are records, with digital IDs, for negatives that were "Not On Shelf" (e.g., dc0934 #'s 75, 76, 77 & 78 - Vietnam Memorial).  Some appear to have been scanned (they have a funny DIGID [e.g., 1.92E6p: SC0264 #16], show data in the LUT fields, and DON'T have a batch # assigned] but are NOT in the "delivered images" database, nor are they on the server.  Assuming we have a negative, these would need to go onto a scan list.  

Because of its size and the difficulty of working with the four Excel tables, it would take a bit of work to come up with a list that clarifies the status of the additional 27,000 records, including eliminating records that have no photos, and creating an "NOS" and "Need to scan" list.  Fortunately, the data available in the JJT database is very rich and a great deal of information can be mined from it.  It would seem that this work would not be a priority until we were planning to gear up the photo scanning again.  Martin's database additions would be based upon the "scanned images" database which ONLY contains those images which have been delivered.  

Let me know what else I should do with this at this point in time, and do we still want to consider finding a place for it to reside on the server to provide additional back-up of the data?  (Approximately 460 MB for all four spreadsheets combined.)        - Kit  

>>> Phil Michel 9/16/2004 8:20:01 AM >>>
It won't translate, and doesn't need to.  LUT stands for look-up table, it's purely binary/math data applied by CONVERT to the files as they are adjusted.  THe OIMGIC fields are checksum signatures for the various stages of the images.
 
You're right about drawings, nothing in CONVERT for those.
 
If other stuff looks okay and the total records for the photos seems about right, then should be okay.  YOu could look at the batch numbers, too, to see if they are all represented.
 
-Phil

>>> Katherine A. Peterson 9/15/2004 6:27:50 PM >>>
Phil,

See attached for notes and a scroll through samples. I'm guessing the gobbtly-gook is CONVERT look-up table stuff that simply might not translate. Would have to look again more closely to try to determine relevance. 

Didn't find any drawings, but they were scanned through a separate system that wasn't integrated into CONVERT so might not have had that kind of data capture. 

Tried unsuccessfully to import and open in ACCESS to see if the gobblty-gook could be translated. (Crashed it 3x at various stages of the import process, never got very far.) - Kit