questions on m.examine.tape and i.tape.other to run on a file!

Michael Shapiro shapiro at zorro.cecer.army.mil
Mon Aug 24 10:18:07 EDT 1992


m.examine.tape is only for 9 track tapes. 9 track tapes how a special
property that m.examine.tape takes advantage of to determine the
"physical" layout of the tape. It will NOT work for a file since the
"physical" structure of a file is hidden by the operating system (and
wouldn't be worth much to you in any case). m.examine.tape is intended
as a first pass to figure out how to extract data from tape if you
don't have the programs that know what the data already looks like.

If you are trying to extract imagery of dem data from data that used to
be on tape but has been put into a file, you will need to know how
that data was structure on the tape BEFORE you extract it to a file.

Hopefully when data developers use STDT as a transfer standard and system
have modules that import STDT files many of these problems will go away
(but not all - that implies a perfect world). STDT should be relatively
media independent so if the data is on 9track or 1/4" cartridge or in a file
it won't matter.

|
|
|	We were trying to run m.examine.tape on a file. It asked for the size of
|buffer. Depending on the size of buffer no. of rows and columns keep changing.
|So, what is the best way to do this?
|
|Size of the TM image file is :  99140608
|If we run m.examine.tape on this file using a buffer size of 32767
|len  :  records
|32767:  1-3025
|20437:3026-3026
|
|Using size of buffer: 40000
|m.examine.tape gives
|len   : records
|40000 : 1-2478
|20608 : 2479-2479
|
|	To extract the same TM image:
|       While running i.tape.other if we select BSQ format (and all
|bands are in a single file)
|is it CORRECT to give 
|
|3026  for total number of records
|32767 for the length of the longest record
|
|       What should I enter for the end row and end column for
|selecting the window to extract ?
|
|	Do I make any sense ?
|	Appreciate your help.
|
|	Laxmi P. Musunur (LP) 
|
|	mlp at forestry.auburn.edu
|
|
|



Michael



More information about the grass-user mailing list