The test in this directory is derived from that in the Berkeley mpeg
package, but we have trimmed it down for the purposes of the Netpbm
package to save space.

First of all, the Berkeley package had Berkeley YUV files for sample input
(in the 'ts' directory).  But we converted them to JPEG to save a great 
deal of space (even when the package is gzipped).

We modified the ts.param file to work with the JPEG files and updated the
expected results file to correspond to the JPEG input (the expected results
are different after the JPEG conversion because JPEG conversion is lossy).

We kept some of the other parameter files from the Berkeley package in
case someone can get some use out of them, but did not update them to
use the JPEG files.  You'll have to do that yourself.  And we removed
the expected results files for them, since you can't generate those
results with the inputs we have supplied.

Note that JPEG input doesn't really exercise the most standard function
of ppmtompeg.  As its name suggests, it's main purpose is to take PPM
frames as input.  You can use jpegtoppm to create PPM input and then
modify the parameter files if you want to test that.  We have not, however
supplied an expected results file for that.

The .mpg files are what ppmtojpeg output as the mpeg movie when we
ran it on the jpeg files in the 'ts' directory.

The .stat files are what ppmtojpeg output when we used the -stat
option in generating the expected results.  As part of your test, use
the -stat option and compare yours to the expected results.