mpeg2encode / mpeg2decode ========================= MPEG-2 Encoder / Decoder, Version 1.2, July 19, 1996 Copyright (c) 1996 MPEG Software Simulation Group E-mail: mssg@mpeg.org (author contact) Web: http://www.mpeg.org/MSSG/ FTP: ftp://ftp.mpeg.org/pub/mpeg/mssg/ Contents: 1. Overview 2. Introduction 3. Contacting the MPEG Software Simulation Group 4. Availability 5. Installation 6. Acknowledgements 7. History of the technical report 1. Overview =========== This directory contains our implementation of an ISO/IEC DIS 13818-2 codec. It converts uncompressed video frames into MPEG-1 and MPEG-2 video coded bitstream sequences, and vice versa. The files mpeg2enc.doc and mpeg2dec.doc in the doc/ directory contain further information about the codec. The directory verify/ contains a small set of verification pictures, a small bitstream, and Unix shell script to automatically test the output of the encoder and decoder. A precompiled version of the programs for Win32s (Windows NT/95) will be made available later date, although it is trivial to make a console application from the encoder and decoder with most Win32s compilers (such as Microsoft Visual C++). Subdirectories src/mpeg2enc and src/mpeg2dec contain the source code for the encoder and decoder, subdirectory par/ contains a couple of example encoder parameter files for 25 and 30 frames/sec MPEG-2 and MPEG-1 video. Summary of changes since July 4, 1994 release: This is only the second official release of our MPEG-2 video software. Only minor bug corrections have been added to the encoder. We still do not implement scalable encoding, as this is mostly useful only for academic research. The decoder has been updated to meet the final MPEG specification, although the old decoder will still reconstruct Main Profile and MPEG-1 bitstreams just fine. The current decoder implements the most important case of Spatial scalability, as well as SNR and Data Partitioning. Temporal scalability is not implemented. 2. Introduction =============== MPEG-2 Video is a generic method for compressed representation of video sequences using a common coding syntax defined in the document ISO/IEC 13818 Part 2 by the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC), in collaboration with the International Telecommunications Union (ITU) as Recommendation H.262. The MPEG-2 concept is similar to MPEG-1, but includes extensions to cover a wider range of applications. The primary application targeted during the MPEG-2 definition process was the all-digital transmission of interlaced broadcast TV quality video at coded bitrates between 4 and 9 Mbit/sec. However, the MPEG-2 syntax has been found to be efficient for other applications such as those at higher bit rates and sample rates (e.g. HDTV). The most significant enhancement over MPEG-1 is the addition of syntax for efficient coding of interlaced video (e.g. 16x8 block sizes for motion compensation, field dct organization, Dual Prime prediction, et al). Several other more subtle enhancements (e.g. 10-bit DCT DC precision, non-linear macroblock scale quantizer, intra VLC tables, improved IDCT mismatch control) were adopted which have a moderate improvement in coding efficiency.... even for progressive video sequences. Other key features of MPEG-2 are the scalable extensions which permit the division of a continuous video signal into two or more coded bitstreams representing the video at different resolutions (spatial scalability), picture quality (SNR scalability and data partioning), or frame rates (temporal scalability). The MPEG Software Simulation Group is currently developing MPEG software with the purpose of providing aid in understanding the various algorithms which comprise an encoder and decoder, and giving a sample implementation based on advanced encoding models. The MPEG-2 software project is an on-going development. Since the current version of the encoder already employs a reasonable (and the most popular) subset of the MPEG-2 signal coding toolkit (MPEG-1 and MPEG-2 Main Profile), and there appears to be sufficient public interest, we have decided to make a public release of the code. The encoder can also be used for generating good quality constant bitrate MPEG-1 sequences and is (to our knowledge) the first publicly available encoder based on the relatively sophisticated TM5 coding model. 3. Contacting the MPEG Software Simulation Group ================================================ We welcome any project-specific questions, comments, suggestions, bug reports etc. They should be sent to the Internet E-mail address: mssg@mpeg.org which automatically forwards to the software authors. 4. Availability =============== The most recent version of the codec source code is available by anonymous ftp from: ftp://ftp.mpeg.org/pub/mpeg/mssg/ The directory contains the following files: mpeg2vidcodec_v12.tar.gz codec source code and documentation mpeg2v12.zip source code and Win32s executables You need gunzip (GNU zip/unzip) to uncompress the .gz and .zip archives. 5. Installation =============== mpeg2decode and mpeg2encode have been compiled and tested on the following platforms: - SUN SPARCstation 10, SunOS 4.1.3, (gcc compiler) - '386-PC, MSDOS 5.0, djgpp v1.11, gcc v2.5.7 and MS Visual C++ 4.0 The source code should compile without major modifications on other 32 bit systems with ANSI C compliant compilers. Systems with 16 bit 'int' variables or segmented memory models are not supported. Please report any modifications you had to apply in order to install the programs on your system to the address mssg@mpeg.org The encoder and decoder are kept in separate sub-directories, src/mpeg2dec contains the decoder, while src/mpeg2enc contains the encoder sources. The following installation procedure applies to both the encoder and the decoder: Step 1: edit Makefile --------------------- You may have to set CC to your C compiler and CFLAGS to the flags required by the compiler. It is sufficient to set these variables in the top directory Makefile. They are propagated to the individual Makefiles of the encoder and decoder. Any other changes have to be applied to the individual Makefiles, however. You can compile the decoder with or without X11 output. Please follow the instructions in the top-level Makefile to activate X Window System support. Step 2: edit src/mpeg2dec/config.h ---------------------------------- In most cases, no modification should be required. If your C library doesn't accept "rb" / "wb" parameters in fopen() (required to disable LF <-> CR/LF conversion on MSDOS systems), change the #defines RB and WB to "r" and "w". Step 3: make ------------ Type 'make' from the top directory (mpeg2). On a PC with DJGPP installed you have to enter 'make pc' in the individual source directories to produce .exe files. Step 4: verification -------------------- In the mpeg/verify directory, you can verify correct compilation of the codec by typing 'make test'. No differences should be reported. The only comparison which is allowed to fail is between test.m2v and new.m2v, caused by floating point accuracy dependencies in the forward DCT. 6. Acknowledgements =================== Authors of the current release are: Stefan Eckart Chad Fogg 420to422, 422to444 scaling filters: Cheung Auyeung Windows 32s port: Sorin Papuc Special thanks are due to - J. Steurer, M. Oepen, IRT (Institut fuer Rundfunktechnik, Muenchen): for contributing motion estimation speed improvements (distance computation short-circuit in conjunction with spiral search, cf. dist1(), fullsearch()) - Tristan Savatier for his help on numerous improvements, suggestions, and features. Numerous users: for providing bug reports and Makefiles 7. History of Technical Report Project ====================================== The Technical Report, a document which primarily consists of a C source code program, was initiated by the MPEG committee to: - Provide an example of MPEG video syntax being intelligently employed to generate good quality video bitstreams. - A reference tool for implementors - Aid in understanding the MPEG specification - decoder which employs full arithmetic accuracy. ---- End of Readme file