The two utility routines here are useful for converting
between the 8-bit/channel sgi rgb format and the hdf format.

----------------------------  SGItoHDF.out  -----------------------------
The routine SGItoHDF.out will read an rgb file with any number
of channels and convert it into an hdf file with the same number
of channels. In doing so, it "reverses" the operations of 
gamma or log scaling, and allows you to convert the 8-bit 
integers into a selected real number range.   

The input parameters for SGItoHDF.out are:

Usage: SGItoHDF [-f firstFrame] [-l lastFrame] [-i increment] 
 [-m minimum] [-M maximum] [-G gamma] [-L] [base_name]

The -G option implies a gamma scaling on the rgb image of "gamma",
and it is reversed.  The -L option implies a log-base-10 scaling
of the rgb image, and that is reversed.  The routine assumes the
rgb files have names of the form

base_nameXXXX.rgb

where XXXX is the four digit frame number.

The actual operation that is performed on each pixel is 

     hdf-value = [ sgi-value * (maximum-minimum) + minimum ]^(1/gamma)

For the -L option, the operation is

    hdf-value = 10^[ sgi-value * (maximum-minimum) + minimum ]

---------------------------  3HDFtoSGI.out  -------------------------

The routine 3HDFtoSGI.out converts a 3-channel hdf file to a
3 channel rgb file.  It has the same arguments as SGItoHDF.out,
but applies them directly instead of the reversed operation in
SGItoHDF.out.

The actual operation performed on each pixels is   

    sgi-value = [ hdf-value^(gamma) - minimum ] / (maximum-minimum)

For the -L option, the operation is

    sgi-value = [ log10(hdf-value) - minimum ] / (maximum-minimum)

In addition to the options also available in SGItoHDF.out, 3HDFtoSGI.out
has a -F option.  The option does two things:

(1) It autoscales each image to the min and max of the data;
(2) It prints to stdout the values of min and max for each channel and 
    frame.

