diff options
Diffstat (limited to 'TAO/orbsvcs/tests/AVStreams/mpeg/INSTALL.UCB')
| -rw-r--r-- | TAO/orbsvcs/tests/AVStreams/mpeg/INSTALL.UCB | 105 |
1 files changed, 0 insertions, 105 deletions
diff --git a/TAO/orbsvcs/tests/AVStreams/mpeg/INSTALL.UCB b/TAO/orbsvcs/tests/AVStreams/mpeg/INSTALL.UCB deleted file mode 100644 index 38a656a538b..00000000000 --- a/TAO/orbsvcs/tests/AVStreams/mpeg/INSTALL.UCB +++ /dev/null @@ -1,105 +0,0 @@ -These are the instructions for compiling mpeg_play, the software MPEG -decoder and player. - -1) Untar the file mpeg_play.tar.Z. This is best done by creating a new -directory for the files involved, moving the tar file into this directory -and invoking the command: - - zcat mpeg_play-2.0.tar.Z | tar xvf - - -2) Create and customize the Makefile. Do this by copying the file -Makefile.proto to Makefile. Then edit Makefile for your particular needs. - -In the makefile the CFLAGS definition is multiply defined for a couple of -different machines, uncomment the one you want to use, or define your own -if necessary. - -The INCLUDEDIR variable should be set to include the paths leading to the -standard header files such as stdio.h as well as to X11/Xlib.h (on most -systems, this is /usr/include). To do this set the INCLUDEDIR variable to --I followed by the pathname. For example, if the path is /usr/local/include, -edit the Makefile to look like this: - - INCLUDEDIR = -I/usr/local/include - -If more than one pathname is necessary, simply prepend -I to each path. -For example if you needed to include /usr/local/include and /usr/X11/include, -the variable definition would look like: - - INCLUDEDIR = -I/usr/local/include -I/usr/X11/include - -Finally, make sure the LIBS variable is set to the path and name of your X11 -library. For example: - - LIBS = /usr/lib/X11/libX11.a - -3) Type make all. - -4) To remove .o files, type make clean - -5) Add yourself to the mailing list by sending mail to - mpeg-list-request@roger-rabbit.cs.berkeley.edu with the subject line - "ADD" (you can delete yourself by sending the subject line "DEL"). - If you are a uunet user, you can use the mail path - 'uunet!ucbvax!roger-rabbit.cs!mpeg-list-request' - - mail mpeg-list-request@roger-rabbit.cs.berkeley.edu - Subject: ADD - ^D - -6) Try it out! You can ftp some sample data files from the same site - you got this player from. Data files usually end in .mpg or .mpeg - The command line for the player is described in the man page, but - is basically: - - mpeg_play [options] [file_name] - -7) If you want the player to collect statistics on size of frames, - macroblocks, time to decode, etc., add the following to the - definition of CFLAGS in the Makefile: - -DANALYSIS - Remove all .o files with "make clean" and remake with "make all" - The player will now print summarized statistics at the end of the - video clip and can be made to print frame by frame statistics with - the use of the -eachstat flag. Read man page for more info. ------------------------------------------------------------------------- - -Using Imake - -We have included an Imakefile for use with Imake. Since we do not use -Imake ourselves and the file was provided by someone else, we can not -give any specific instructions on how to use it. Please, consult -someone more experienced with Imake. -Thanks. - ------------------------------------------------------------------------- -NOTES - -It seems that much of the time is spent converting the 24 bit MPEG image -to an 8 bit color space. This process is called "dithering". We've included -several dithering algorithms. Read the man pages for more instructions. - -The data files available are produced by XING. These images are -usually small (~160X120). XING data does not take advantage -of P or B frames (ie, frames with motion compensation). The data is simply -a series of I frames. Performance of the player on XING data is -significantly lower (half or less) of the performance when motion compensated -MPEG data is decoded. - -Reporting bugs: - If you find any bugs in this software, please send them to - mpeg-bugs@roger-rabbit.cs.berkeley.edu. Since this software - is unsupported, we make no guarantees about how long it will - take to fix the bug, or if it will be fixed at all. Bug fixes - will be cheerfully accepted. Please include as much detailed - information as possible, including: - - 1) the version number of the program you are using (cf. VERSION) - 2) the data file that caused the bug (if possible) - 3) the OS version and machine type you ran the program on - 4) the compiler used to compile the program - - - - - |