Cell file size

[Tom Hawkins]

I am having a difficult time understanding why the cell files I create in 
GRASS are not the size I am led to believe they should be.  

I have a SPOT panchromatic image that I loaded into GRASS.  I didn't use a 
GRASS program to do this.  I put he file under /cell, then created a /cellhd 
file using all the required data from the SPOT header file.  Then I ran 
r.support and am able to d.rast the image and create postscript files that 
plot out fine.  The SPOT file size is equal to the # of rows times the 
number of columns (1 byte per cell). 

I set g.region so that the entire image fit inside the region comfortably.  
I made sure there was no mask set.  I then ran r.resample to create a new 
cell file.  

The new cell file is smaller than what I expected.  The region was set to 
have 5754 rows and 4613 columns, which calculate out to 26,543,202.  The 
actual flie size of this resampled file is 25,789,652.  

What am I doing wrong?  My main goal is to mask out a particular area of the 
image (a 7 1/2 minute quad) and create a new cell file (a BIL file) that can 
be used in another program.  The other program will take a BIL file, meaning 
the rows time columns must equal the file size.  

Any help and guidance would be greatly appreciated.