tm subscene on cd-rom with 3 bands only, extraction with i.tape.other, BAD TAPE READ

[Alexander Brauns]

we have recently acquired a set of landsat-tm subscenes (peten, northern
guatemala) with only the bands 345 (budget!) to perform a change
analysis (95-98) on protected areas, I was not able to extract the
scenes with i.tape.other (grass4.2.1).

the scenes are on cd-rom, we copied them to harddisk
the device (here a path pointing to the .bil file) is apparently accepted,
after filling in the dialogs provided by i.tape.other, the extraction
process starts  and produces the following output:

	advancing to row 1
	extracting 0%
	ERROR bad tape read (row1 band1)
	    raster files will be incomplete
	   
	creating support files ... (etc)

the support files are created correctly, the cell files are small and
contain 0 values (they display as black image);
all the parameters requested by i.tape.other have been provided
according to the image header file, the only doubt left is on the
"blocking factor of data in the tapes", which I assumed to be 1; 
"lenght in bytes of longest record" was filled in with the amount of
columns=pixels, as one pixel should contain one byte;
the source file is OK, we had no problems extracting in IDRISI;
what strikes me as strange is that m.examine.tape does not work, it does
not return any information and creates an empty file, I suspect a
problem with reading the original file because it is not on tape,
however, postings on the grass lists recommend the same procedure for
SPOT imagery on CD-ROM;
i.tape.tm does definitly not work, it does not accept the tape device
name, I assume because it looks directly for 7 bands, which are not
present.

some hints would be greatly appreciated;

yours
alexander brauns
asesor técnico conap-proselva
conappro at quetzal.net