RISC DISC 1/RISC_DISC_1.iso/usefulinfo/faq/mpegfaq

Internet Message Format | 1994-04-03 | 123.5 KB

Path: cix.compulink.co.uk!uknet!pipex!uunet!munnari.oz.au!sol.ccs.deakin.edu.au!nomad
From: [e-mail] (Damon Atkins)
Newsgroups: comp.sys.acorn
Subject: LAST REPOST: THE MPEG-FAQ Version 2.0 - Part 1
Date: 8 Oct 1993 03:19:35 GMT
Organization: Deakin University, Victoria, Australia
Lines: 1453
Message-ID: <292m87$[e-mail]>
References: <28u8bj$[e-mail]
NNTP-Posting-Host: eros.ccs.deakin.edu.au
Keywords: MPEG, FAQ

In article <28u8bj$[e-mail] [e-mail] (Frank Gadegast) writes:


        ======================================================
        THE MPEG-FAQ               [Version 2.0 - 9. May 1993]
        ======================================================
        PHADE SOFTWARE Leibnizstr. 30, 1000 Berlin 12, GERMANY
        Inh. Frank Gadegast            Fon/Fax: [phone]

        [e-mail]


===========================================================================

This is my summary about MPEG.

It's the third publication of this file. Lots of errors have been removed,
and lots of information has been added (which has surely brought other
errors with it, see Murphy's Law).

This third addition is VERY different to the previous ones.

First:  Some sections have been removed, because the are old or there was
        nothing changing. So if you are NOT familiar with the theme, get
        the the last MPEG-FAQ (Version 1.1) via ftp from

        host: ftp.cs.tu-berlin.de (130.149.17.7)
        file: /pub/msdos/windows3/graphics/mpegfa11.zip

        This new FAQ will be there soon too, as 'mpegfa20.zip'.

Second: The E-MAIL-ORDER-SERVICE has been removed, because it was causing
        to much traffic on my account. The damon HAS been changed, it will
        send a message back, saying, that the service has been removed.
        Right now, there are enough placed where to get the desired files
        from (via Internet, E-Mails to others, ftp, mail-access etc.)

Third:  The people where more interested to get the complete archives.
        Therefore the TRAIL-PACK-Service is still running. I'm still
        collecting EVERY info, video, sound or program. Get the Trail-Pack !

You should read carefully through this FAQ this time, cause lots of new
information is hidden in all the sections. F.e. news about Dos, Amiga-,
Atari-, OS/2-, Windows-, Windows-NT, VMS- and Mac-Players arrived !!!


This summary is devided in 12 parts:

 I    |  WHAT IS MPEG ?
 II   |  PROFESSIONAL SOFTWARE
 III  |  PUBLIC-DOMAIN SOFTWARE OR SHAREWARE
 IV   |  MPEG-RELATED HARDWARE
 V    |  MAILBOX-ACCESS
 VI   |  FTP-ACCESS (PD)
 VII  |  MAIL-ORDER
 VIII |  RETRIEVED MAIL OR ARTICLES
 IX   |  ADDITIONAL INFORMATION
 X    |  WHERE TO FIND MORE INFOS
 XI   |  NEWS
 XII  |  QUESTIONS

I add my comments in brackets [], lines (---- or ====) seperate the
chapters.

Please try and find out more information yourself. I had enough to do by
getting and preparing this information. And only bother me with file-
request if its not possible for you to get it somewhere else !!!

If you want to contribute to this FAQ in any way, please email me
(probably by replying to this posting). My email address is:

  [e-mail]

Or send any additional information via fax or e-mail. The fax is only
reachable between Mo.-Fr. from 10.00-13.00 and from 15.00-18.30 german
time.

    Phade (Frank Gadegast)


DISCLAIMER: I HAVE NOTHING TO DO WITH THE NAMED COMPANIES, NO BUSINESS,
            IT'S JUST MY PERSONAL INTERESTED. THESE COMPANIES ARE NAMED,
            BECAUSE, THEY ARE THE FIRST BRINGING MULTIMEDIA TO THE PC-
            WORLD. SURE I MAKE ADVERTS FOR THEM WITH THIS FAQ, BUT HOPE-
            FULLY YOU, AS A READER OF THIS FAQ, WILL FORCE THEM TO PRODUCE
            MORE AND BETTER PRODUCTS.


===========================================================================
 I | WHAT IS MPEG ?
===================

From comp.compression Mon Oct 19 15:38:38 1992
Sender: [e-mail]
Author: Mark Adler <[e-mail]>

[71] Introduction to MPEG (long)
       What is MPEG?
       Does it have anything to do with JPEG?
       Then what's JBIG and MHEG?
       What has MPEG accomplished?
       So how does MPEG I work?
       What about the audio compression?
       So how much does it compress?
       What's phase II?
       When will all this be finished?
       How do I join MPEG?
       How do I get the documents, like the MPEG I draft?

[ There is no newer version of this part so far. Whoever wants to update ]
[ this description, should do the job and send it over.                  ]

------------------------------------------------------------------------------

Subject: [71] Introduction to MPEG (long)


Written by Mark Adler <[e-mail]>.

Q. What is MPEG?
A. MPEG is a group of people that meet under ISO (the International
   Standards Organization) to generate standards for digital video
   (sequences of images in time) and audio compression.  In particular,
   they define a compressed bit stream, which implicitly defines a
   decompressor.  However, the compression algorithms are up to the
   individual manufacturers, and that is where proprietary advantage
   is obtained within the scope of a publicly available international
   standard.  MPEG meets roughly four times a year for roughly a week
   each time.  In between meetings, a great deal of work is done by
   the members, so it doesn't all happen at the meetings.  The work
   is organized and planned at the meetings.

Q. So what does MPEG stand for?
A. Moving Pictures Experts Group.

Q. Does it have anything to do with JPEG?
A. Well, it sounds the same, and they are part of the same subcommittee
   of ISO along with JBIG and MHEG, and they usually meet at the same
   place at the same time.  However, they are different sets of people
   with few or no common individual members, and they have different
   charters and requirements.  JPEG is for still image compression.

Q. Then what's JBIG and MHEG?
A. Sorry I mentioned them. Ok, I'll simply say that JBIG is for binary
   image compression (like faxes), and MHEG is for multi-media data
   standards (like integrating stills, video, audio, text, etc.).
   For an introduction to JBIG, see question 74 below.

Q. Ok, I'll stick to MPEG.  What has MPEG accomplished?
A. So far (as of January 1992), they have completed the "Committee
   Draft" of MPEG phase I, colloquially called MPEG I.  It defines
   a bit stream for compressed video and audio optimized to fit into
   a bandwidth (data rate) of 1.5 Mbits/s.  This rate is special
   because it is the data rate of (uncompressed) audio CD's and DAT's.
   The draft is in three parts, video, audio, and systems, where the
   last part gives the integration of the audio and video streams
   with the proper timestamping to allow synchronization of the two.
   They have also gotten well into MPEG phase II, whose task is to
   define a bitstream for video and audio coded at around 3 to 10
   Mbits/s.

Q. So how does MPEG I work?
A. First off, it starts with a relatively low resolution video
   sequence (possibly decimated from the original) of about 352 by
   240 frames by 30 frames/s (US--different numbers for Europe),
   but original high (CD) quality audio.  The images are in color,
   but converted to YUV space, and the two chrominance channels
   (U and V) are decimated further to 176 by 120 pixels.  It turns
   out that you can get away with a lot less resolution in those
   channels and not notice it, at least in "natural" (not computer
   generated) images.

   The basic scheme is to predict motion from frame to frame in the
   temporal direction, and then to use DCT's (discrete cosine
   transforms) to organize the redundancy in the spatial directions.
   The DCT's are done on 8x8 blocks, and the motion prediction is
   done in the luminance (Y) channel on 16x16 blocks.  In other words,
   given the 16x16 block in the current frame that you are trying to
   code, you look for a close match to that block in a previous or
   future frame (there are backward prediction modes where later
   frames are sent first to allow interpolating between frames).
   The DCT coefficients (of either the actual data, or the difference
   between this block and the close match) are "quantized", which
   means that you divide them by some value to drop bits off the
   bottom end.  Hopefully, many of the coefficients will then end up
   being zero.  The quantization can change for every "macroblock"
   (a macroblock is 16x16 of Y and the corresponding 8x8's in both
   U and V).  The results of all of this, which include the DCT
   coefficients, the motion vectors, and the quantization parameters
   (and other stuff) is Huffman coded using fixed tables.  The DCT
   coefficients have a special Huffman table that is "two-dimensional"
   in that one code specifies a run-length of zeros and the non-zero
   value that ended the run.  Also, the motion vectors and the DC
   DCT components are DPCM (subtracted from the last one) coded.

Q. So is each frame predicted from the last frame?
A. No.  The scheme is a little more complicated than that.  There are
   three types of coded frames.  There are "I" or intra frames.  They
   are simply a frame coded as a still image, not using any past
   history.  You have to start somewhere.  Then there are "P" or
   predicted frames.  They are predicted from the most recently
   reconstructed I or P frame.  (I'm describing this from the point
   of view of the decompressor.)  Each macroblock in a P frame can
   either come with a vector and difference DCT coefficients for a
   close match in the last I or P, or it can just be "intra" coded
   (like in the I frames) if there was no good match.

   Lastly, there are "B" or bidirectional frames.  They are predicted
   from the closest two I or P frames, one in the past and one in the
   future.  You search for matching blocks in those frames, and try
   three different things to see which works best.  (Now I have the
   point of view of the compressor, just to confuse you.)  You try using
   the forward vector, the backward vector, and you try averaging the
   two blocks from the future and past frames, and subtracting that from
   the block being coded.  If none of those work well, you can intra-
   code the block.

   The sequence of decoded frames usually goes like:

   IBBPBBPBBPBBIBBPBBPB...

   Where there are 12 frames from I to I (for US and Japan anyway.)
   This is based on a random access requirement that you need a
   starting point at least once every 0.4 seconds or so.  The ratio
   of P's to B's is based on experience.

   Of course, for the decoder to work, you have to send that first
   P *before* the first two B's, so the compressed data stream ends
   up looking like:

   0xx312645...

   where those are frame numbers.  xx might be nothing (if this is
   the true starting point), or it might be the B's of frames -2 and
   -1 if we're in the middle of the stream somewhere.

   You have to decode the I, then decode the P, keep both of those
   in memory, and then decode the two B's.  You probably display the
   I while you're decoding the P, and display the B's as you're
   decoding them, and then display the P as you're decoding the next
   P, and so on.

Q. You've got to be kidding.
A. No, really!

Q. Hmm.  Where did they get 352x240?
A. That derives from the CCIR-601 digital television standard which
   is used by professional digital video equipment.  It is (in the US)
   720 by 243 by 60 fields (not frames) per second, where the fields
   are interlaced when displayed.  (It is important to note though
   that fields are actually acquired and displayed a 60th of a second
   apart.)  The chrominance channels are 360 by 243 by 60 fields a
   second, again interlaced.  This degree of chrominance decimation
   (2:1 in the horizontal direction) is called 4:2:2.  The source
   input format for MPEG I, called SIF, is CCIR-601 decimated by 2:1
   in the horizontal direction, 2:1 in the time direction, and an
   additional 2:1 in the chrominance vertical direction.  And some
   lines are cut off to make sure things divide by 8 or 16 where
   needed.

Q. What if I'm in Europe?
A. For 50 Hz display standards (PAL, SECAM) change the number of lines
   in a field from 243 or 240 to 288, and change the display rate to
   50 fields/s or 25 frames/s.  Similarly, change the 120 lines in
   the decimated chrominance channels to 144 lines.  Since 288*50 is
   exactly equal to 240*60, the two formats have the same source data
   rate.

Q. You didn't mention anything about the audio compression.
A. Oh, right.  Well, I don't know as much about the audio compression.
   Basically they use very carefully developed psychoacoustic models
   derived from experiments with the best obtainable listeners to
   pick out pieces of the sound that you can't hear.  There are what
   are called "masking" effects where, for example, a large component
   at one frequency will prevent you from hearing lower energy parts
   at nearby frequencies, where the relative energy vs. frequency
   that is masked is described by some empirical curve.  There are
   similar temporal masking effects, as well as some more complicated
   interactions where a temporal effect can unmask a frequency, and
   vice-versa.

   The sound is broken up into spectral chunks with a hybrid scheme
   that combines sine transforms with subband transforms, and the
   psychoacoustic model written in terms of those chunks.  Whatever
   can be removed or reduced in precision is, and the remainder is
   sent.  It's a little more complicated than that, since the bits
   have to be allocated across the bands.  And, of course, what is
   sent is entropy coded.

Q. So how much does it compress?
A. As I mentioned before, audio CD data rates are about 1.5 Mbits/s.
   You can compress the same stereo program down to 256 Kbits/s with
   no loss in discernable quality.  (So they say.  For the most part
   it's true, but every once in a while a weird thing might happen
   that you'll notice.  However the effect is very small, and it takes
   a listener trained to notice these particular types of effects.)
   That's about 6:1 compression.  So, a CD MPEG I stream would have
   about 1.25 MBits/s left for video.  The number I usually see though
   is 1.15 MBits/s (maybe you need the rest for the system data
   stream).  You can then calculate the video compression ratio from
   the numbers here to be about 26:1.  If you step back and think
   about that, it's little short of a miracle.  Of course, it's lossy
   compression, but it can be pretty hard sometimes to see the loss,
   if you're comparing the SIF original to the SIF decompressed.  There
   is, however, a very noticeable loss if you're coming from CCIR-601
   and have to decimate to SIF, but that's another matter.  I'm not
   counting that in the 26:1.

   The standard also provides for other bit rates ranging from 32Kbits/s
   for a single channel, up to 448 Kbits/s for stereo.

Q. What's phase II?
A. As I said, there is a considerable loss of quality in going from
   CCIR-601 to SIF resolution.  For entertainment video, it's simply
   not acceptable.  You want to use more bits and code all or almost
   all the CCIR-601 data.  From subjective testing at the Japan
   meeting in November 1991, it seems that 4 MBits/s can give very
   good quality compared to the original CCIR-601 material.  The
   objective of phase II is to define a bit stream optimized for these
   resolutions and bit rates.

Q. Why not just scale up what you're doing with MPEG I?
A. The main difficulty is the interlacing.  The simplest way to extend
   MPEG I to interlaced material is to put the fields together into
   frames (720x486x30/s).  This results in bad motion artifacts that
   stem from the fact that moving objects are in different places
   in the two fields, and so don't line up in the frames.  Compressing
   and decompressing without taking that into account somehow tends to
   muddle the objects in the two different fields.

   The other thing you might try is to code the even and odd field
   streams separately.  This avoids the motion artifacts, but as you
   might imagine, doesn't get very good compression since you are not
   using the redundancy between the even and odd fields where there
   is not much motion (which is typically most of image).

   Or you can code it as a single stream of fields.  Or you can
   interpolate lines.  Or, etc. etc.  There are many things you can
   try, and the point of MPEG II is to figure out what works well.
   MPEG II is not limited to consider only derivations of MPEG I.
   There were several non-MPEG I-like schemes in the competition in
   November, and some aspects of those algorithms may or may not
   make it into the final standard for entertainment video compression.

Q. So what works?
A. Basically, derivations of MPEG I worked quite well, with one that
   used wavelet subband coding instead of DCT's that also worked very
   well.  Also among the worked-very-well's was a scheme that did not
   use B frames at all, just I and P's.  All of them, except maybe one,
   did some sort of adaptive frame/field coding, where a decision is
   made on a macroblock basis as to whether to code that one as one
   frame macroblock or as two field macroblocks.  Some other aspects
   are how to code I-frames--some suggest predicting the even field
   from the odd field.  Or you can predict evens from evens and odds
   or odds from evens and odds or any field from any other field, etc.

Q. So what works?
A. Ok, we're not really sure what works best yet.  The next step is
   to define a "test model" to start from, that incorporates most of
   the salient features of the worked-very-well proposals in a
   simple way.  Then experiments will be done on that test model,
   making a mod at a time, and seeing what makes it better and what
   makes it worse.  Example experiments are, B's or no B's, DCT vs.
   wavelets, various field prediction modes, etc.  The requirements,
   such as implementation cost, quality, random access, etc. will all
   feed into this process as well.

Q. When will all this be finished?
A. I don't know.  I'd have to hope in about a year or less.

Q. How do I join MPEG?
A. You don't join MPEG.  You have to participate in ISO as part of a
   national delegation.  How you get to be part of the national
   delegation is up to each nation.  I only know the U.S., where you
   have to attend the corresponding ANSI meetings to be able to
   attend the ISO meetings.  Your company or institution has to be
   willing to sink some bucks into travel since, naturally, these
   meetings are held all over the world.  (For example, Paris,
   Santa Clara, Kurihama Japan, Singapore, Haifa Israel, Rio de
   Janeiro, London, etc.)

Q. Well, then how do I get the documents, like the MPEG I draft?
A. MPEG is a draft ISO standard. It's exact name is ISO CD 11172.
   The draft consists of three parts: System, Video, and Audio. The
   System part (11172-1) deals with synchronization and multiplexing
   of audio-visual information, while the Video (11172-2) and Audio
   part (11172-3) address the video and the audio compression techniques
   respectively.

   You may order it from your national standards body (e.g. ANSI in
   the USA) or buy it from companies like
     OMNICOM
     phone [phone]
     FAX [phone]


===========================================================================
 II.1 | PROFESSIONAL SOFTWARE
=============================

Xing Technology Corporation
PO Box 950                                Voice: 805-473-0145
456 Carpenter Canyon                      FAX:   [phone]
Arroyo Grande, CA 93420

Xing products include:

    MPEG Motion video capture/encode and decode.
    JPEG Photo image encode and decode.
    Video capture boards and associated software for both JPEG and MPEG.

    Microsoft Windows Applications, DOS Applications,
        and Software Developers Kits are available for JPEG and MPEG.

---------------------------------------------------------------------------

Check out the latest in Frame Grabber technology, the

      PC-Hurricane,

a realtime  true color frame grabber, which can digitize about 500 frames
in realtime (25 frames/sec) into Extended Memory (32 MBytes).
So it gives you 20 seconds of full-motion video on the PC.

These 320 frames can be saved with one command to the harddrive and can
then be processed to a MPEG file with just one other command.
You can then join several 20 seconds MPEG clips together to a whole
MPEG movie with the MPEG utilities.

PC-Hurricane, only available from Ingenierbuero Gatz & Hartmann, GERMANY.


-------------------------------------------------------------------------------
 II.2 |
-------

Ingenieurbuero Gatz & Hartmann,

Fehrbelliner Str. 32, 1000 Berlin 20, GERMANY

Tel: [phone] or 030-375 55 68
FAX: [phone] or 030-375 56 55

email to: [e-mail] (Stefan Hartmann)


The MPEG Encoder is available starting from 349.-DM incl. VAT.

---------------------------------------------------------------------------

PC-Hurricane, only available from Ingenierbuero Gatz & Hartmann.
It is available for a price of 699.-DM inclusive 14 %  VAT.

Here comes the technical info about the new moviegrabber(tm)
PC-Hurricane(tm):

It is a hicolor color YUV (4:1:1) realtime video movie digitizer which can
store 25 frames/sec(PAL) or 30 frames/sec(NTSC)  into Expanded Memory. It
has support for ET4000 Hicolor and true color boards.

The software is a userfriendly DOS software, which displays the incoming
video signal in almost realtime (about 5-10 frames/sec) on a ET4000 Hicolor
board in 32768 colors/pixel. This is okay for video purposes, cause the
noise inside a video movie almost does already the dithering to a true
color quality (simualar to 24 bit quality).

If you don't have a Hicolor card inside your PC, then there is also a 256
color version driver, which displays the video signal in almost realtime
in 3-3-2 color quantisation (still the TGA pics are saved as 16 bit
Hicolor/pixel, so quality from the saved output is the same as in the
Hicolor software driver).

If the sequence arrives, which you wanna digitize, just hit the space key.
Then it is digitized into the Expanded Memory in realtime with 25 or 30
frames/sec maximum rate. This can be adjusted to 12.5 or 8 or 4 frames/sec!
During this storing process, you will not see any signal on your VGA
monitor, but if the RAM is filled with e.g.396 frames at 320x240, then you
can already watch the digitized movie play of the Expanded Memory!

It plays in slow motion the YUV coded fields to the SVGA card via the ISA
bus. This gives a real slow motion Hicolor digital movie out of the RAM !

So if you have 32 MBytes of Ram on your motherboard, you can be happy.
Then you can store about 16 seconds of realtime MPEG resolution 320x240
true color video with one shot ! With the enclosed software you can save
this with one command to single TGA pics onto the harddisk or save it as a
complete Hicolor movie (uncompressed, big file size).

PC-Hurricane also digitizes the resolutions: 384x288, 360x240, 320x240 and
320x200. So max. resolution is 1 PAL field (384x288). Field digitisation
is preferred, cause it gives no Interlace artifacts, due to motion...

So if you want to make a FLI file out of the 16 seconds sequence, no
problem, just start DTA17e (Dave's Targa Animator) and it produces a FLI
animation file out of the single TGA pics. Also, if you have the Xing
Technology software only MPEG encoder, it will be possible to compress
this about 400 single TARGA(tm) pics with just one second command to a
MPEG movie ! So you have in about 5 minutes of work a ready to display
MPEG movie !

PC-Hurricane has NTSC/PAL switchable. So it can also digitize NTSC video.
It has Video(FBAS) and Y/C(SVHS/Hi8) inputs.

---------------------------------------------------------------------------

MPEG 2.0 for windows3.x is now available !

It is the digital Video player via a software only solution ! It displays
in a 320x240 window under win3.x a realtime decompressed digital video !
Decompression is done only by software and it reaches 30 frames/sec on a
486 PC ! The new version has a very enhanced picture quality, because the
compression rate with the encoder can now be adjusted ! The very new thing
is the WAV-Sound support ! So if you have a soundcard inside your PC [or
the speaker-drv installed !], you will have a real video-clip with
accompagning sound !

They currently sell 3 demo disks with the full featured Player, version
2.0 and lots of animations on the disks.

It is available for 39.-DM over here in Germany, which is 26 US$.

[ It does no sounds over the speaker-drv (works only with sound-card), ]
[ does now recognize Windows 3.1, and does OLE, but only if you buy    ]
[ the MCI-driver. Works for nearly every card under 640x480x256. Works ]
[ with ET-4000 under 1024x768x256 (only with small fonts) too.         ]

[        ATTENTION !!! THIS ONE IS NOT UP TO DATE ANYMORE !!!          ]
[======================================================================]
[ This player is the one, you can get now for free from Xing's         ]
[ BBS or find on several ftp-servers. This player was an offer to get  ]
[ this player even before it was available for the public-domain. Thnx ]
[ to Stefan !                                                          ]

[ Here the description of the PD-Player :                              ]

MFW.EXE Version 2.0

1.1 General Description

MFW can be started from the DOS command line or as a "run" command from
within Windows.  From the DOS prompt, enter WIN MFW.  From the Windows FILE
RUN dialog, enter MFW.  Using either of the two command line methods,
additional command parameters can be passed to the program for customizing
its startup.

Valid command line parameters will take precedence over the customization
parameters saved in the MFW.INI file.  Invalid command line parameters will
be ignored as if missing altogether.  The MFW.INI file provides default
startup values for each missing command parameter except the optional file
name.  The MFW.INI file is updated at user discretion from menu options
selected while running MFW.

[      ATTENTION ! SEVERAL BUGS ARE KNOWN FOR THIS PLAYER !!              ]
[=========================================================================]

[ The player doesn't detect where it resided, by checking the argv[0]     ]
[ argument (complete pathname !).                                         ]

[ So, if you start the player via double-click on a .mpg-File from the    ]
[ FILEMAN.EXE (or similar), the player doesn't find it's MFW.INI-File. It ]
[ can't read its system then, and the hole thing is fu**ed !!             ]

[ Or the player doesn't find his .INI-File, thinks that this was the      ]
[ first time it was started and:                                          ]

[   LEAVES A NEW MFW.INI-File in the current directory !!!                ]

[ Not in the one the player is, no ! In the current one. So my movies are ]
[ in g:\mpeg\movies, the player is somewhere under c:\windows.31\mpeg, I  ]
[ double-click on the .mpg-Movie and get a new MFW.INI in g:\mpeg\movie ! ]

[-------------------------------------------------------------------------]

[ Now the second 'maybe not so bad' bug:

[ The description says, to put the corresponding DLL-Files in the SYSTEM- ]
[ directory.  That's NOT neccessary, cause the DLL-loading-API-function   ]
[ will look through the path-variable and try to load the DLL's from      ]
[ there.                                                                  ]

---------------------------------------------------------------------------

BTW, the encoder still sells for 349.-DM and the MCI-driver for 199.-DM

[ The MCI-driver is nice, because it allows you to include movies in      ]
[ other documents. But it includes only the MPLAYER.EXE-icon in the       ]
[ document (not the first picture of the movie), the movie runs at        ]
[ whatever position (not where the icon is !), when you double-click it.  ]

[ Xing should have a close look at Microsoft's AVI-driver ;o) (but there  ]
[ movies are incredible slow and small, compared to MPEG  :o(             ]

---------------------------------------------------------------------------

24 bit to TV converter box which is a box connected outside of the PC to
the SVGA card via its VGA connector. There must be a TSR program run in the
background to change automatically the scanning frequencies of the SVGA
board. This works also very well with win3.1. It supports up to 800x600x
65K colors (only PAL) and 640x480 with 24 bits color (true color) on the
Genoa 7900 ET4000 SVGA board. It has  Video (FBAS) and Y/C (SVHS/HI8)
outputs.

There is also a version of this box with additional RGB output, so you get
the best picture quality on a TV-set with SCART - RGB-inputs..

The VGA Monitor is also connected to the 24 Bit to TV converter box, so if
you don't start the driver software, your system is like normal: You see
the picture on your SVGA monitor. The DC-power comes from the keyboard
connector, so you don't have to fiddle around with AC-DC-aptor power
supplies !

Just don't start the driver software, if you don't need video out ! The box
doesn't interfere with your daily work. Just forget it behind you PC. If
you need video out, just start the 2 KBytes driver software and then you
can toggle between video out and VGA monitor display with a hot key !

The driver software is adapted to SVGA FLIC-files from Autodesk Animator
PRO to give a 640x288 x256 colors Fullscreen-Overscan display on PAL TV-
sets. This works by using double scanning with 50 Hz Non-Interlace !
Every scanline from the 288 is scanned twice, so it gives 640x576
Fullscreen display ! (with no flicker)

This way you can record your favourite SVGA animations to tape without
seeing any black borders !

There is also a NTSC version available, which works only up to 640x480 with
16, 256 ,32K , 65 K and 24 bits colors ! On NTSC model the 640x480 mode is
already Fullscreen !

It also works with the new Hicolor FLI player, so you can play and record
hicolor animation directly from a RAMdisk to VCR tape. You don't need NO
single frame accurate expensive VCR anymore ! Cause it works also with
win3.1, you can record all your daily windows application work to video
tape !

If you have a presentation program like Microsoft Powerpoint, no problem !
Just record your presentation to video tape to show it via video cassette
to everyone, who has a VCR !

In windows the 24 bit to TV-box works in 50 or 60 Hz Interlace to display
the 640x480 or 800x600 (only PAL) scanlines. This gives a little flicker on
Menu-buttons, which have only very small horizontal lines with big contrast
in color. So use an appropriate color palette under windows to reduce
flicker..

With the Y/C output (SVHS/Hi8) you almost get SVGA monitor quality on your
video tape !

---------------------------------------------------------------------------

VGA2TV adaptor. This is an internal board which is hooked inside the PC to
the feature connector of the VGA card. So the VGA or SVGA board must have a
feature connector ! The VGA2TV samples the contens of the VGA RAM into its
own dual ported RAM and gives it out with the Video scanning frequencies to
its Video (FBAS) or Y/C output. It also has a Genlock possibility. So it
can be used to do titling in front of a holiday movie video tape running in
the background of the computer graphics titles !

Input of the external video source could also be FBAS or Y/C. This unit is
called M2150 for the PAL version. M2100 is the NTSC version.

---------------------------------------------------------------------------

VGA2TV PRO with integrated Flicker fixer. This is the new release of the
VGA2TV board. It has also a flicker filter for making the Interlace flicker
much less visible. It also has OverScan/Underscan feature and the position
of the screen on the TV set could be controlled via software. This board is
of course more expensive than the older VGA2TV, but still either cards are
sold.  The new one is called M2200 .

Both max resolutions to convert with these VGA2TV boards are 640x480 with
256 colors ! It has the advantage, that it does not need any software !
Just plug and it plays !

---------------------------------------------------------------------------

Productlist from Gatz & Hartmann
================================

Genoa 7900 SVGA 24 bit true color Multimedia Graphik-Karte.... 299.-DM
24 bit to TV Konverter Box, passend zur Genoa 7900 Karte.......349.-DM
Genoa 7900 und 24 bit to TV Box, im Bundle.....................629.-DM
24 bit to TV Box mit zusaetzlichem RGB-Video Ausgang auf Scart.549.-DM
VGA2TV Genoa Genlock-Karte,intern, Computergrafik ueber Video..998.-DM
VGA2TV PRO, Profi-Version mit Flicker-fixer,etc...............1599.-DM
PC-Titler Standard,Video Vertitelungs-Software fuer VGA-Karten.349.-DM
PC-Titler Deluxe, Profi-Version vom PC Titler, mehr Effekte....649.-DM
The Video Workshop, PC gesteuerter Video-Schnittplatz.........1599.-DM
The Video Workshop Lite, Druckerport gesteuerter Video-Schnitt.999.-DM
MPEG Encoder VTM, digitales Video per Software,fuer TGA-pics...349,-DM
MPEG MCI-win3.x-driver,Treiber zum Einbinden von MPEG-clips ...199.-DM
MPEG 2.0 Demo, 3 Disketten mit dem neuen WAV-support player ....39.-DM
PC-Hurricane,25 Bilder/sec realtime MPEG true color Digitizer..699.-DM
FG-02 , Hicolor DTP Framegrabber, 512x512 pixel Videodigitizer.499.-DM
VT-Express,sehr schneller JPEG-Bild viewer, ca. 1.2 sec/Bild...349.-DM
Hicolor(32768Farben) Animations-Software f. TGA-Animationen ...149.-DM

[ Looks like, if I don't have to translate this ;o)                  ]

[ Well, I know, looks like advert, but they are the only ones, that  ]
[ produce MPEG-Hard- and Software for this price. Hereby I declare,  ]
[ everybody can send me their lists, too. I will publish them in the ]
[ next FAQ. Current Dollar-DeutschMark-course is: 1$ = 1,67 DM !!!   ]


===========================================================================
 III | PUBLIC-DOMAIN-SOFTWARE OR SHAREWARE
==========================================


---------------------------------------------------------------------------
 III.1 | DOS
------------

[ This is the NEW Xing-Dos-Player ! ]

The MPEG-Player 'MPLAY.EXE' from Xing Technologies is included
in the 'MPEGXING.LZH'-file.

Hi, this is the new Full-Screen DOS only MPEG player from Xing Technology.

Use it under DOS:  mpeg name.mpg

This will give a fullscreen 320x200 display with 256 colors. You can play
with it the Xing MPEG files which are normally 320x240 screen size.

With this shareware player you have 40 scanlines less, so upgrade to the
upcoming XMode DOS player which will have 240 scanlines and will also
support WAV audio with it.

For more info just contact Xing Technology under:

805-473-0145 (voice)  or
805-473-0147 (fax)

(in the USA)

Best regards, Stefan Hartmann
email to: [e-mail]

---------------------------------------------------------------------------

There is another player called PLAYMPEG.ZIP including two executables
MPEG128.EXE and MPEG250.EXE (for 128 and 250 colors) by:

PLAYMPG v. 0.9 - Jan. 1993 - by Giampiero Caprino, Olivetti, Italy.
This is a pre-release version being distributed for evaluation

The modified berkeley-source-code is avilable as MPEGSRC.ZIP.

[ The user-interface is more comfortable than the MPEG.EXE one, but ]
[ these players here are terrible slow. Anyway, it shows how easy   ]
[ you can adapt the berkeley-code (see below) and port it to other  ]
[ systems. And it works on ALL VGA-cards !!!                        ]

[ Hopefully, Rowe, Patel and Smith include this additional code in  ]
[ the next version of their MPEG-Player to get a system-undependent ]
[ distribution.                                                     ]

---------------------------------------------------------------------------

[ A new implementation of the berkeley-code for DOS. This code is   ]
[ available for AT286/386 and the files are named MPEG286.ZIP and   ]
[ MPEG386.ZIP .                                                     ]

MPEG translator and player V0.00

This is the 286 version of the MPEG traslation and playback routines.
I would like to thank Giampero Caprino for providing the VGA routines and
the authors of the Mpeg standard Unix distribution for most of this code.

To play back an animation you first must translate it to a raw data file.
To do this type

   mpeg <dither-options> input-file output-file

You will see it counting off the frames as it runs. When it is finished
type

   playback <delay-value> input-file

Where input-file to playback corresponds to the output-file of mpeg.exe.
The delay value defaults to 0, but this is probably too fast.
On a 486sx with no delay, I get just over 20 frames per second.
Try values between 0 and 200.

Note that several dithering options are not supported: 2x2, hybrid2,
and gray. Perhaps I will fix this in the future.

Also please note that if the MPEG animation loops, the translator will
happily write a file as big as your disk drive. Hit CTRL-C to stop it.

This program is freeware.
I have included the original Unix man page and docs for the MPEG player.

Greg Ennis
[e-mail]

---------------------------------------------------------------------------

From: [e-mail] (Stefan Eckart)
Subject: dmpeg10.zip info: Another DOS MPEG decoder/player posted
Date: Wed, 14 Apr 1993 08:53:14 GMT

I have posted a DOS MPEG decoder/player to alt.binaries.pictures.utilities.

Here is a short description and some technical information, taken from the
accompanying documentation:

                              DMPEG V1.0
                       Public Domain MPEG decoder
                           by Stefan Eckart

0. Features
===========

DMPEG/DMPLAY is another MPEG decoder/player for the PC:


 - decodes (nearly) the full MPEG video standard
   (I,P,B frames, frame size up to at least 352x240 supported)

 - saves decoded sequence in 8 or 24bit raw file for later display

 - optional on-screen display during decoding (requires VGA)

 - several dithering options: ordered dither, Floyd-Steinberg, grayscale

 - color-space selection

 - runs under DOS, 640KB RAM, no MS-Windows required

 - very compact (small code / small data models, 16 bit arithmetic)

 - real time display of the raw file by a separate player for
   VGA and many Super-VGAs

...

4. Technical information
========================

The player is a rather straightforward implementation of the MPEG spec [1].
The IDCT is based on the Chen-Wang 13 multiplication algorithm [2]
(not quite the optimum, I know). Blocks with not more than eight non-zero
coefficients use a non-separated direct multiply-accumulate 2D-IDCT
(sounds great, doesn't it?), which turned out to be faster than a 'fast'
algorithm in this (quite common) case. Dithering is pretty standard. Main
difference to the Berkeley decoder (except for the fewer number of supported
algorithms) is the use of 256 instead of 128 colors, the (default) option to
use a restricted color-space and the implementation of a color saturation
dominant ordered dither. This leads to a significantly superior quality of
the dithered image (I claim, judge yourself).

Restricted color-space means that the U and V components are clipped to
+/-0.25 (instead of +/-0.5) and the display color-space points are distributed
over this restricted space. Since the distance between color-space points
is thus reduced by a factor of two, the color resolution is doubled at the
expense of not being able to represent fully saturated colors.

Saturation dominant ordered dither is a method by which a color, lying
somewhere between the points of the display color space, is approximated
by primarily alternating between two points of constant hue instead of
constant saturation. This yields subjectivly better quality due to the
lower sensitivity of the human viewing system to saturation changes than
to hue changes (the same reasoning as used by the PAL TV standard to improve
on NTSC). The improvement is particularly visible in dark brown or redish
areas.

Stefan Eckart, [e-mail]


---------------------------------------------------------------------------
 III.2 | WINDOWS
----------------

The MPEG-Player 'MPEGXING.LZH' from Xing Technologies.

[ Good player ! Works for nearly every card under 640x480x256   ]
[ Works with ET-4000 under 1024x768x256 (with small fonts) too. ]

---------------------------------------------------------------------------

The Genoa product presentation MPEG demo. Its available via
Ingenieurbuero Gatz & Hartmann,

This is what you need to play it:

  A fast 486 ( at least 33 Mhz) with 16 MBytes of RAM.
  Also it is a ET4000 board required. Local-Bus is also okay...

  [ Also you need about 20MB free disk space. 10 for the file ]
  [ and 10 for the SMARTDRV-Cache. Didn't try it on the PC,   ]
  [ but on the Sun's. It's a animated product info, not a     ]
  [ movie; not worth to bother !!                             ]

---------------------------------------------------------------------------

MPEGPLAY V1.0 (c) 1993 Michael Simmons

[ Hup ! Anotherone, that thinks, he's alone ... I proclaim, that this ]
[ player should be called MPEGW32 from now on !!!                     ]

This is Release Version 1.0 of an MPEG player for WIN32(tm) and WIN32s(tm).

This player can play standard mpeg files that include P and B frame
encoding, and large 354x288 movie files.

It has several display options including mono,gray scale,color dither and
Full color (for Hicolor graphic cards).

To install the player under Windows 3.1(tm), Unzip the file disk1.zip
to a floppy disk. Then, run the setup.exe file from Windows.

To install the player under Windows NT(tm) copy the files mpegplay.exe and
mpegplay.hlp to a common directory. Then, create a new program item for the
mpegplay.exe file via the File New option of the Program Manager.

The working directory must be the same as the program path. Otherwise
the Online help file will not be found (see known bugs section at the end
of this document.)

Read the Disclaimer in the online Help before loading any mpeg movie files.

This program is SHAREWARE.

To receive the latest version of the Player send $25 US and a list of
suggestions and/or bugs to:

        Michael Simmons
        34 Shillington Way
        THORNLIE WA 6108
        AUSTRALIA

DISTRIBUTION:

This File must not be separated from the rest of this archive.

Due to licensing conditions of the WIN32s(tm) System this archive can only be
Redistributed in the following ways:
(1) Archive site to End user.
(2) Archive site to Archive site.
The following means of redistribution are not permitted:
(1) End user to End user.
(2) End user to Archive site.

Redistribution from Archive site to Archive site may only be performed by
the operators of those sites.
An Archive site is taken to be any large collection of software which is
operated by a person or group of persons for the primary purpose of
redistributing that software.
An End user is taken to be the person or group of persons who use this
software.

Known Bugs:
(1) Help file must be in the working directory when the player is started.
(2) The Mono Dither is not working properly.
(3) The 2x2 Color Dither has patches of incorrect color.
(4) Sometimes the player fails to display Full Color 354x288 images after
    a previous movie has been Opened and then Closed.
(5) The Player is slower than is possiable.
(6) Bug/feature The Player runs slow when ever the mouse is moving.

Windows NT, Win32s, Windows 3.1 are trademarks of the Microsoft Corporation.

ACKNOWLEDGEMENTS:

This code was derived from the U.C. Berkeley MPEG Player (version 2.0)
developed by L.A. Rowe, K. Patel, and B. Smith ([e-mail]).
That code included the following copyright:

  Copyright (c) 1992 The Regents of the University of California.
  All rights reserved.


---------------------------------------------------------------------------
 III.3 | WINDOWS-NT
-------------------

From: [e-mail] (Michael Simmons - mgmt_staff)
Subject: MPEG PLAYER for NT/WIN32s uploaded to cica
Date: 16 Mar 1993 02:49:03 GMT

I have uploaded my port (alpha version) to Windows NT and Win32s of
the berkeley Public Domain MPEG player to ftp.cica.indiana.edu
It is in the /pub/pc/win3/uploads directory.File name mpegnt.zip

Below is the contents of the mpegnt.txt file.

This archive contains my port (alpha version) to Windows NT(tm) and Win32s
of the berkeley Public Domain MPEG player.

WARNING this is code is very buggy so use it at your own risk.
Read all of this file before running the player.

I suggest that you close any running applications before running this player.

To run this program under Windows 3.1 you will need a copy the the Win32s
runtime this can be found in the Spice4?.zip files in the /pub/pc/win3/nt
directory on ftp.cica.indiana.edu.
It is also include on the October 92 Win32 SDK CD-ROM.

For Win32s+Window3.1 users I also suggest you restart windows and/or reboot
your machine after running the player.

This player can Play both Xing(tm) 160x120 MPEG movies and standard
320x240 encoded movies.

Note at present only the 24bit full color ditherer is implemented. This means
that dithering the image down to 256 or 16 colors is done via the Windows
DIB's driver. For best results you need either a 24bit or 15bit (hicolor)
graphics card. Using Hicolor mode gives much better images than the Xing(tm)
Player.
At present I am working on getting the other ditherers working.

This player has been run on the following two machines:

(1) OPTI 486DX33 with 8Mb,Trident 1MB Video card in 640x480x16 Mode,120MB
    Hard disk and running Windows NT(tm).

(2) SMI 486DX33 with 8Mb,ATI 2MB ULTRA + in 800x600x65K Mode, 200MB Hard disk
    and runing Win32s and Windows 3.1.

It may not run on a machine with only 4Mb.

To install the player copy the files to your hard disk then create a new
program item by selecting the Program Manager File/New menu item.

There are two versions of the player included mpegplay.exe and mpegply2.exe.
mpegply2 is the same as mpegplay.exe but attempt to completly hog the cpu.
The machine will not respond to any input while it is playing movies.

To Play a movie perform the following:
(1) Start the player either from the NT command prompt or via the
    Program Manager.
(2) Select the Open Menu item from the File Menu.
(3) type "*.mpg" into the file name dialog box.
(4) select the file you wish to play.
(5) Select Play from the Movie Menu
(6) The movie will now play.
(7) if you are running the mpegplay.exe player you may select Stop
    from the Movie Menu and then Open another movie for playing.
    Note - The player will Exit when/if it gets to the end of the Movie.


The "Actual Size" and "Stretch to Window" options in the Display Menu
determine whether the movie images are displayed as actual size (centered
in the middle of the windows client area) or scaled to fill the windows
entire client area.
The Player will also play as an Icon if Iconified.

This player is very easy to cash.

Some known Bugs are
(1) image is horizontally flipped.(Hey at least its not upside down!).
(2) Player will cash if another movie is opened while the current movie
    is playing.
(3) Stretch to window and Iconic play back modes are not functioning
    under win32s.
(4) The common dialog box file selection filters are not working.
(5) Fails to free memory after exit.

The next release of this player should be much more stable.

Please send suggestions and comments to [e-mail]

Windows NT(tm) is a trademark of Microsoft Corporation.
Xing(tm) is a trademark of Xing.

---------------------------------------------------------------------------
 III.4 | OS/2
-------------

From: [e-mail] (Tom Hatton)
Subject: Re: OS/2 MPEG player???
Date: 19 Apr 93 20:14:46 GMT

There appears to be an OS/2 MPEG player; haven't used it, but the
hobbes (ftp-os2.nmsu.edu) listing shows the following:

mpegplay.zip      97028  Full-screen 320x200 MPEG animation player

in pub/os2/2.x/graphics.

[ Would be nice, if somebody could test this, and post some results. ]


---------------------------------------------------------------------------
 III.5 | X-WINDOWS and Unix
---------------------------

The Berkeley Plateau Research Group is happy to announce the release of
Version 2.0 of its software-only MPEG decoder. The player is available via
anonymous ftp from toe.cs.berkeley.edu (128.32.149.117) in
/pub/multimedia/mpeg/mpeg_play-2.0.tar.Z. You'll find many sample MPEG
streams in the subdirectory movies.

Changes from v1.2 include:

o Fixed green artifact bug.
o Fixed sequence end code bug.
o Many bug fixes.
o Performance tweaks.


                  MPEG Video Software Decoder
                  (Version 2.0; Jan 27, 1993)

        Lawrence A. Rowe, Ketan Patel, and Brian Smith
 Computer Science Division-EECS, Univ. of Calif. at Berkeley

This directory contains a public domain MPEG video software
decoder. The decoder is implemented as a library that will
take a video stream and display it in an X window on an 8, 24
or 32 bit deep display.  The main routine is supplied to
demonstrate the use of the decoder library. Several dithering
algorithms are supplied based on the Floyd-Steinberg, ordered
dither, and half-toning algorithms that tradeoff quality and
performance. Neither the library nor the main routine handle
real-time synchronization or audio streams.

The decoder implements the standard described in the Committee
Draft ISO/IEC CD 11172 dated December 6, 1991 which is
sometimes refered to as "Paris Format." The code has been
compiled and tested on the following platforms:

 HP PA-RISC (HP/UX 8.X, X11R4) (i.e., HP 9000/7XX and 9000/3XX)
 Sun Sparc (SunOS 4.X, X11R5)
 DECstation 5000 and Alpha
 IBM RS6000
 Silicon Graphics Indigo
 MIPS RISC/os 4.51
 Sequent Symmetry
 Sony NEWS
 and more than we can list here.

If you decide to port the code to a new architecture, please let
us know so that we can incorporate the changes into our sources.

This directory contains everything required to build and
display video. We have included source code, a makefile, an Imakefile,
installation instructions, and a man page. Data files can
be obtained from the same ftp site this was located in.
See the INSTALL file for instructions on how to
compile and run the decoder.

The data files were produced by XING. XING data does not take
advantage of P or B frames (ie, frames with motion compensation).
Performance of the player on XING data is significantly slower
(half or less) than the performance when motion compensated MPEG
data is decoded. We are very interested in running the software
on other MPEG streams.  Please contact us if you have a stream
that does not decode correctly. Also, please send us new streams
produced by others that do utilize P and B frames.

NOTE: One particular XING data file: raiders.mpg, is not a
valid MPEG stream since it does not contain a sequence
header.

We have established several mailing lists for messages about
the decoder:

[e-mail]
   General information on the decoder for everyone interested
   should be sent to this list.  This should become active after
   11/20/92

[e-mail]
   Requests to join or leave the list should be sent to this
   address. The subject line should contain the single word
   ADD or DELETE.

[e-mail]
   Problems, questions, or patches should be sent to this address.

Our future plans include porting the decoder to run on other
platforms, integrating it into a video playback system that
supports real-time synchronization and audio streams, and
further experiments to improve the performance of the
decoder. Vendors or other organizations interested in supporting
this research or discussing other aspects of this project should
contact Larry Rowe at [e-mail].

We also plan on producing an MPEG encoder. The encoder will NOT be
a real time digitizer, but will be intended for offline processing
of video data.

ACKNOWLEDGEMENTS:
    We gratefully thank Hewlett-Packard, Fujitsu, the Semiconductor
    Research Corporation for financial support.

    We also want to thank the following people for their help:

    Tom Lane of the Independent JPEG Group provided us with
        the basic inverse DCT code used by our player.
        ([e-mail])

    Reid Judd of Sun Microsystems provided advice and assistance.
    Todd Brunhoff of NVR provided advise and assistance.
    Toshihiko Kawai of Sony provided advise and assistance.

[ Brilliant !!! With a bit a power from a Sun or something like ]
[ this, MPEG can be real fun. There is no real user-interface,  ]
[ but the quality is first class !! Very easy to compile !!!    ]

[ Version 1.2 is just relased. Now with Imakefile and lots of   ]
[ bug-fixes. Executable is much smaller. Still gets some colors ]
[ wrong (with some movies).                                     ]

[ Version 2.0 is release and fixes this color- and other bugs.  ]

[ Running as well on :
[ PC 386/486     ISC 2.2.1/3.0, SCO, 386BSD  X11R3/R4/R5        ]
[ NeXT           NeXTStep                    NeXT Window, X11R5 ]
[ Mac's          AUX (with gcc1.37)          X11-lib            ]

---------------------------------------------------------------------------

[ Brand new. The codec ! Meaning a Public-Domain-Encoder-Kit for Unix ! ]

From: [e-mail] (Michael Simmons - mgmt_staff)
Subject: Standford MPEG codec
Date: Thu, 25 Feb 1993 16:07:18 +0800 (WST)


                 MPEG, CCITT H.261 (P*64), JPEG
  Image and Image sequence compression/decompression C software engines.


The Portable Video Research Group at Stanford have developed
image/image sequence compression and decompression engines (codecs)
for MPEG, CCITT H.261, and JPEG. The primary goal of these codecs is
to provide the functionality - these codecs are not optimized for
speed, rather completeness, and some of the code is kludgey.

Development of MPEG, P64, and JPEG engines is not the primary goal of
the Portable Video Research Group.  Our research is focused on
software and hardware for portable wireless digital video
communication.  For more information about current research, please
send e-mail to Professor Teresa Meng at [e-mail].

COMMENTS/DISCLAIMERS:

This code has been compiled on the Sun Sparc and DECstation UNIX
machines; some code has been further checked on the HP workstations.

For comments, bugs, and other mail relating to the source code, we
appreciate any comments. The code author can be reached at Andy C.
Hung at [e-mail].  The standard public domain disclaimer
applies: Caveat Emptor - no guarantee on accuracy or software support.

References related to these codecs should NOT use any author's name,
or refer to Stanford University.  Rather the Portable Video Research
Group or the acronym (PVRG) should be used, such as PVRG-MPEG,
PVRG-P64, PVRG-JPEG.

CODEC DESCRIPTION:

I) PVRG-MPEG CODEC: (havefun.stanford.edu:pub/mpeg/MPEGv1.0.tar.Z)

This public domain video encoder and decoder was generated according
to the Santa Clara August 1991 format.  It has been tested
successfully with decoders using the Paris December 1991 format. The
codec is capable of encoding all MPEG types of frames. The algorithms
for rate control, buffer-constrained encoding, and quantization
decisions are similar, but not identical, to that of the (simulation
model 1-3) MPEG document.  The rate control used is a simple
proportional Q-stepsize/Buffer loop that works though not very well -
better rate-control is the essence for good quality buffer-constrained
MPEG encoding.  Verification of the buffering is possible so as to
provide streams for real-time decoders.

The MPEG codec performs compression and decompression on raw raster
scanned YUV files. The companion display program for the X window
system is described in section IV) below.  A manual of approximately
50 pages describes the program's use.

There are also MPEG compressed files from the table tennis sequence in
tennis.mpg and the flower garden sequence in flowg.mpg.

This codec was recently tested with the MPEG decoder of the Berkeley
Plateau Research group. If what you want is decoding and X display,
then you might want to look into their faster public domain MPEG
decoder/viewer. The Berkeley player is available via anonymous ftp
from toe.cs.berkeley.edu (128.32.149.117) in
/pub/multimedia/mpeg/mpeg-1.2.tar.Z.

II) PVRG-P64 CODEC: (havefun.stanford.edu:pub/p64/P64v1.0.tar.Z)

This public domain video encoder and decoder is based on the CCITT
H.261 specification.  Some encoding algorithms are based on the RM 8
encoder.  The codec has been tested against itself, though we were
unable to test it against the INRIA encoder because apparently INRIA
also interleaves/packetizes the audio and video. We would appreciate
anyone having p64 video test sequences to let know.  Like the MPEG
codec, it supports all the encoding and decoding modes, and has
provisions for buffer-constrained encoding, so it can produce streams
for real-time decoders.

The H.261 codec takes the similar YUV raster scanned files as the MPEG
codec, and performs compression and decompresion on raster scanned YUV
files.  It can take standard CIF or NTSC-CIF files. The display of
these programs is described in section IV) below.  A manual of
approximately 50 pages describes its use.

There are also P64 compressed files from the table tennis sequence in
table.p64 and the flower garden sequence in flowg.p64.

III) PVRG-JPEG CODEC: (havefun.stanford.edu:pub/jpeg/JPEGv1.0.tar.Z)

This public domain image encoder and decoder is based on the JPEG
Committee Draft.  It supports all of the baseline for encoding and
decoding.  The JPEG encoder is flexible in the variety of output
possible.  It also supports lossless coding, though not as speedy as
we would like.  The manual is approximately 50 pages long which
describes its use.  The display program for JFIF-style (YUV) files is
described in section IV) below.  The JFIF style is not a requirement
for this codec - it can compress and decompress CMYK, RGB, RGBalpha,
and other formats - this codec may be helpful if you wish to extract
information from non-JFIF encoded JPEG files.

This codec has been tested on publicly available JPEG data.  For
general purpose X display, you might want to try the program "xv"
(version 2.0 or greater).  The JPEG engine of the program "xv" is
based on the free, portable C code for JPEG compression available from
the Independent JPEG Group.  (anonymous login - ftp.uu.net (137.39.1.9
or 192.48.96.9) /graphics/jpeg/jpegsrc.v4.tar.Z).

IV) X VIEWER: (anonymous login-  havefun.stanford.edu:pub/cv/CVv1.0.tar.Z)

This viewer allows the user to look at image or image sequences
generated through the codecs described above. These image or image
sequences are in the YUV colorspace and may be 4:1:1 (CIF style) or
2:1:1 (CCIR-601 style) or 1:1:1 (non-decimated style). A short manual
of approximately 2 pages describes its use.

ACKNOWLEDGEMENTS:

Funded by the Defense Advanced Research Projects Agency.

I am especially grateful to Hewlett Packard and Storm Technology for
their financial support during the earlier stages of codec
development.  Any errors in the code and documentation are my own.
The following people are acknowledged for their advice and assistance.
Thanks, one and all.

    The Portable Video Research Group at Stanford: Teresa Meng,
    Peter Black, Ben Gordon, Sheila Hemami, Wee-Chiew Tan, Eli Tsern.

    Adriaan Ligtenberg of Storm Technology.
    Jeanne Wiseman, Andrew Fitzhugh, Gregory Yovanof of Hewlett Packard.
    Eric Hamilton and Jean-Georges Fritsch of C-Cube Microsystems.

    Lawrence Rowe of the Berkeley Plateau Research Group.


---------------------------------------------------------------------------
 III.6 | MAC
------------

From: [e-mail] (Rainer Menes)
Subject: MPEG player for the Mac!!!
Date: Thu, 11 Mar 1993 15:58:53 GMT

Dear mac users,

To bring the romuors to an end. To day I announce the MPEG player for the
Mac. The distribution is based on the X-Window player 2.0 from Berkley. From
a mac users view I would rate the program not to be a real mac application,
but it does its duty. The packeged includes two application one to view
MPEGs on a Mac in a window, and another to convert MPEGs to PICTs. These
PICTs can be use to convert a MPEG film to a quicktime film. This way isn't
a hit but it works. On a Quadra 700 you get up to 5 frames a sec for some
MPEG films. So don't be dispointed when your old MacII only gets 1 frame/
sec. or so. MPEG is quit heavy stuff for realtime MPEG I have seen that you
need around 1 Gops (Giga Operations per Sec.) to decode and 4-5 Gops for
encoding. This shows you why MPEG algorithm in software is so slow. One
remark for PC users the XING mpeg player for Windos isn't a real mpeg
player, it has a lot of limitations which the MPEG player for the didn't
have, and this makes this thing alot faster. The Mac player should  be able
to view any standard MPEG file, even that one produced by the sortly
available hardware from c-cube.

Now I have talk a lot, here is the ftp site you will find my MPEG player.

  ftp suniams1.statistik.tu-muenchen.de (131.159.64.1)

  login: anonymous
  password: your mailaddress

  cd /pub/mac

  filename: mpeg_mac_0.15.sit.hqx (BinHex4.0 format and Stuffit 1.51 format)

and alot of example files.

If you have the old version of MPEG player for the Mac 0.1 try to get the
new one 0.15. This version should be compatible also to Macs without FPU,
but a 68020 or up is requered. (MacII, MacIIcx, MacIIx, MacIIci, MacIIsi,
SE-30, MacIIfx, Clasic II, LC I II III, Quadra 700, 800, 900, 950, Centris
610 and 650 with or without FPU, Color Clasic, and Power Books 145,160,165,
170,180). You should be able to use the player and converter on any screen
from 538x384 to 1180x890 and form 1-Bit to 32-Bit per pixel.

To avoid problems please read the Readme file which is include in the sit
file. This file gives you some tips to work around some pitfalls with my
player.

Now have fun with the software,

Rainer

email: [e-mail]

P.S: If you have problems please mail me, but sometimes it may take some
     time to get help, because today we got a big project and I will have
     very little time.

[ See his other mail in the section Chapter VIII | RETRIEVED MAIL too. ]


---------------------------------------------------------------------------
 III.7 | ATARI
--------------

From [e-mail] Tue Apr 13 17:56:15 1993
Date: 13 Apr 93 10:40:34 EDT
From: Brainstorm <[e-mail]>


Bonjour!,

  For technical points about our MPEG decoder: It's a SOFTWARE ONLY decoder,
currently running on any ATARI FALCON computers (not on the TT, because it
uses the MOTOROLA DSP56001).

  Video output modes:

- B/W, using a 4x4 bayer dithering matrix.
- 4 greyscales, using a 4x4 bayer dithering matrix.
- 16 greyscales, using a 2x2 bayer dithering matrix.
- ATARI 15 bits true color (no dithering).
We have to write the 256 colors and the 24 bits true color modes.

  Decoder's MPEG compatibility:

It currently only handles I Frames. We have found some MPEG files with P and
B frames, so it shouldn't a problem to implement P and B decompression. We
don't handle MPEG sound, because it's currently impossible to find MPEG files
with MPEG sound. The only card which seems to encode MPEG sound is from
Optibase, and I ask them to give us a MPEG file with sound, but there were




Path: cix.compulink.co.uk!uknet!pipex!uunet!munnari.oz.au!sol.ccs.deakin.edu.au!nomad
From: [e-mail] (Damon Atkins)
Newsgroups: comp.sys.acorn
Subject: LAST REPOST: THE MPEG-FAQ Version 2.0 - Part 2
Date: 8 Oct 1993 03:20:46 GMT
Organization: Deakin University, Victoria, Australia
Lines: 1563
Message-ID: <292mae$[e-mail]>
References: <28u8eb$[e-mail]
NNTP-Posting-Host: eros.ccs.deakin.edu.au
Keywords: MPEG, FAQ

In article <28u8eb$[e-mail] [e-mail] (Frank Gadegast) writes:


not very 'interested' in helping us...

Decoder speed:

It depends a lot of the MPEG file and display mode. Here are some benchmark
with 4 known MPEG files:

File        Size (Bytes)    Time (Frames/Sec)    Video Mode

F16.MPG        94959        12.15            320x200, 4 Col. (TV)
F16.MPG                11.25            320x200, 16 Col. (TV)
F16.MPG                 9.10            320x200, T.C. (TV)

MJACKSON.MPG    724567        13.82            320x200, 4 Col. (TV)
MJACKSON.MPG            13.19            320x200, 16 Col. (TV)
MJACKSON.MPG            10.40            320x200, T.C. (TV)

ROM.MPG        250937         9.25            320x200, 4 Col. (TV)
ROM.MPG                 8.81            320x200, 16 Col. (TV)
ROM.MPG                 8.02            320x200, T.C. (TV)

TAHITI.MPG    303547        11.87            320x200, 4 Col. (TV)
TAHITI.MPG            11.28            320x200, 16 Col. (TV)
TAHITI.MPG             9.45            320x200, T.C. (TV)

  Decoder Size:

Currently, it's about 22Kb. P and B frames management shouldn't take more
than 10Kb.

Next Changes:

1: Speed: We expect to be 2 Frames/Sec faster in true color mode, and
   1 Frame/Sec faster in other modes.

2: P and B frames compatibility.

Decoder's Availability:

Brainstorm is a small company, dedicated in writing software. We don't want
to sell the MPEG decoder by ourself. That's why we are currently looking for
a company interested in buying and distributing it. So, I can't tell you
when it will be available, and what will be it's price.

Brainstorm:
19 bis, rue de Cotte
75012 Paris FRANCE
tel +(331) 44670809
fax +(331) 44670811

We can be reached at:
- [e-mail]
or, maybe,
- [e-mail] ??? (we didn't test it)

---------------------------------------------------------------------------

From [e-mail] Fri Apr 30 16:08:25 1993
Date: 29 Apr 93 10:50:06 EDT
From: Brainstorm <[e-mail]>

Brainstorm MPEG changes:

Things are moving a lot! Last days, we have been working on speed
improvements on our MPEG decoder. Here are the new benchmarks:

File        Size (Bytes)    Av. Time (Frames/Sec)    Video Mode

BIRDISBA.MPG    502473        15.24            320x200, 4 Col. (TV)
BIRDISBA.MPG            14.75            320x200, 16 Col. (TV)
BIRDISBA.MPG            11.51            320x200, T.C. (TV)

[ stuff deleted ]

The decoder is now 32 Kb.

As you can see, the speed depends a lot of the screen resolution. There are
two main reasons:

1 - The Falcon video chip is connected to the MC68030 bus. It means that the
video ram can be everywhere on the falcon RAM (on a PC, the video ram is on
the video card). So, if you are using your computer on a nice video mode
(i.e. true color), a lot of bus bandwidth is used by the video chip, and the
68030 is slower.

2 - In true color mode, the MPEG movie is displayed in color, and we perform
6 8x8 DCTs, and the colorspace conversion. In greyscale mode, we only perform
4 8x8 DCTs, and there is no colorspace conversion (we use the Y channel).

  One of our biggest problem in this decoder is that the Falcon is a '2 chips'
computer: the MC68030 at 16 MHz, and the DSP56001 at 32 MHz. We are using
both chips (the MC68030 for the Huffman decoder, the DSP56001 for everything
else). The speed of the decoder is the speed of the slowest chip. Till last
week, the slowest was the MC68030, but we have improved a lot the Huffman
decoder. So, the bottleneck is now the DSP56001. Our next big job is the DSP
code improvement.

Best regards, Raphael (Brainstorm)


---------------------------------------------------------------------------
 III.7 | DATA
-------------

Several data-files (.mpg) are known. See the list below:

---- Utilities (Players): ----

MPEGDOS.ZIP     22183  11-16-92
MPEGWIN.ZIP    462053   1-18-93
MPEG-2.0.TAR.F  99921   2-15-93
MPEG286  ZIP   115831   2-18-93
MPEG386  ZIP   114109   2-18-93
MPEGPLAY ZIP    97351   3-20-93 [ that's known as PLAYMPEG.ZIP too ! ]
MPEGNT   ZIP   140428   3-17-93

[ These are files from the Stanford-codec's ]

CHANGES             1737 20-03-93
CVv1.1.tar.z       47849 20-03-93
JPEGv1.1.tar.z    172657 20-03-93
MPEGv1.1.tar.z    250732 20-03-93
P64v1.1.tar.z     202249 20-03-93


-------- MPEG-MOVIES ---------

2JAEGER  MPG    70015   3-21-93  10:03a
2JETS    MPG   416159  11-23-92   2:08p
BIRDISBA MPG   502473   7-23-92   1:36a
BIRDSHOW MPG   180963   6-04-92   1:01a
BIRDWALK MPG   206417   6-04-92  12:48a
EGGCLOCK MPG   157341  11-15-92   6:36p
F16      MPG    94959   6-04-92   1:28a
FIMPSY   MPG   281960   3-20-93  11:21a
FIMPSY50 MPG   240029   3-20-93  11:21a
FIREFING MPG    37700   4-23-93   1:53p
FLIGHT   MPG       24  11-21-92  12:37p
FRISCO   MPG    84552   1-07-93   6:31p
HULAHOOP MPG   114148   1-18-93   7:54p
IICM     MPG  1679360   2-24-93   3:15p
JETS     MPG   479434   3-20-93  11:21a
JOEL     MPG   285388   3-20-93  11:21a
MICKY    MPG    53411   3-20-93  11:22a
MJ       MPG   619275   2-11-93   2:07a
MJACKSON MPG   784741   2-05-93   2:54p
MODEL1   MPG   678804   3-20-93  11:22a
MOGLIE   MPG   292665   6-25-92   6:28p
MONKEY   MPG   169142   4-21-93   5:22p
MOONFLAG MPG   118218   4-21-93   5:23p
MOONFLY  MPG   109622   4-21-93   5:25p
MPGGENOA MPG  8241296  11-08-92  10:24p
QUME     MPG   364256   3-20-93  11:22a
RAIDERS  MPG   978660  12-21-92   2:33p
ROCKET   MPG   104987   4-21-93   5:25p
ROM      MPG   250937  11-24-92  10:13a
SUKHOI   MPG   140288   3-20-93  11:23a
TEST10   MPG    32176   2-18-93   1:59p
TEST2    MPG   102188   2-18-93   1:59p
TEST30   MPG    14098   2-18-93   1:59p
TEST4    MPG    63550   2-18-93   1:59p
TEST6    MPG    47399   2-18-93   1:59p
TEST8    MPG    37823   2-18-93   1:59p
TOASTER  MPG    78093   4-21-93   5:25p
XTITLE   MPG     2738  11-15-92   6:36p

------ XING-AUDIO-FILES -------

2JAEGER  WAV    51244   9-03-92   6:27p
2JETS    WAV    59580  11-23-92   2:08p
BIRDISBA WAV   145452   9-03-92   6:30p
BIRDWALK WAV    49196   9-03-92   6:51p
EGGCLOCK WAV    63572   8-25-92   9:58a
JETS     WAV    73772   9-03-92   6:03p
MICKY    WAV    20566   9-03-92   5:37p
MJ       WAV   286784   2-11-93   2:10a
MODEL1   WAV   522290   2-11-93  12:27a
MPGGENOA WAV  1520314  11-08-92   7:31p
RAIDERS  WAV  1301458   3-25-93   5:30p


------ BIG-MPEG-MOVIES -------

BICYCLE  MPG   718897   3-19-93   2:02p
BIKE     MPG   642590   3-20-93  12:38p
BUS      MPG   718464   3-19-93   2:02p
CANYON   MPG  1744060   3-19-93  12:54p
CT       MPG   119040   4-21-93   5:26p
FLOWERS  MPG   690185   3-20-93  12:38p
HULA_2   MPG   148076   2-15-93   1:28p
MOBILE   MPG   573440   3-17-93   3:29p
SHORT    MPG    83522   3-24-93   8:30p
TENNIS   MPG  1246001   3-20-93  11:08a
WATERSKI MPG   417792   3-20-93  11:11a

[ These BIG mpeg-movies are ONLY playable with the berkeley-decoder ! ]
[ Xing does not support this format so far.                           ]


---- MPEG-DOCUMENTS ----

IMAGE-PS F     725803   1-14-93  ( )   [ overview about image-processing ]
MM93-PS.F      242555  11-01-93  ( )   [ overview about MPEG decoding    ]
MPEGSRC.ZIP     92780   1-14-93  ( )   [ the source for MPEG128.EXE      ]
MPLAY-PS.F     205580   1-12-93  ( )   [ overview about a Media Player   ]

[ These are files from the Stanford-codec's. They include several ]
[ Postscript-Pictures (describing the encoding process). The real ]
[ documents are included in the src-archives of the codec's.      ]

JPEGDOCv1.1.tar.z  49873 20-03-93
MPEGDOCv1.1.tar.z  47645 20-03-93
P64DOCv1.1.tar.z   47497 20-03-93


---------------------------------------------------------------------------
 III.8.1 | INFOS about Movies
-----------------------------

The movie SAMPLE.MPG is now know as EGGCLOCK.MPG (surely the .WAV changed
too).

---------------------------------------------------------------------------

IICM.MPG was produced at the Institute for Information Processing and Computer
Supported New Media (IICM) of the Graz University of Technology. It features a
fly-by through (parts of) the Mandelbrot set through a terrain model of the
surroundings of Graz with a corresponding satellite image used as texture map,
and the 'Uhrturm' (clocktower), which is the landmark of Graz and the logo of
the institute.
                            Kein Wunder!
It was produced using Wavefront's Advanced Visualizer on Silicon Graphics 4D/35
and converted to MPEG using Xing's encoder software for Personal Computers.
Distribute freely!

Frank M. Kappe                                     [e-mail]
Institute for Information Processing and Computer Supported New Media (IICM)
                  Graz University of Technology, Austria
Voice: ++43/316/832551-22                               Fax: ++43/316/824394

---------------------------------------------------------------------------

From: [e-mail] (Stefan Hartmann (Behse))
Subject: MPEG CD-ROM, how to submit animations.
Date: 13 Feb 1993 13:07:16 GMT

Hi,

well I'm looking for animations "freaks" who want to see their work
published on a CD-Rom.

We will put together a CD-Rom with many MPEG animations and hope to find
some people who will donate their FLI or FLIC animations to convert to MPEG
animation format.
We will do the conversion, so you just have to donate your FLI or FLC file.

What will it bring to you, if you donate your animation ?

1. You will be included with your full adress in the doc files, so
companies could contact you, if they wish to have some new art from you.

2. This CD-Rom will sell worldwide, so you might get known a little bit
more than now...

3. You will receive 5 free copies of this CD-Rom, so you can sell it to
your friends and earn some money from it.

4.The CD-Rom price will be not higher than about 39 US$, so you will get
about 195 US$ for your work. You can get additional copies of this CD-Rom
for to sell it to more friends at a very low price.

5. You will receive no royalities from this CD-Rom, but it sells at this
low price which just covers the work for producing and distributing this
disc worldwide.

6. Don't miss your chance to be included on the world's first MPEG
CD-ROM with your best animation.

7. The best animation will win a prize !

So I hope this will convince you to contact me and that you would like to
include your animation on our CD-Rom.
If you still have any questions, please send me email.

Best regards, Stefan Hartmann, c/o Gatz & Hartmann
email to:  [e-mail]

Questions and answers:
======================


I have 24-bit stills that create an animation.  I do not know how to create
MPEG animations.  The format is 24-bit Amiga IFF.  Can we do business?


Yes we can ! Just convert your single pictures via e.g. Art Department to
JPEG pics and tell me on what FTP server you might put them up, so that I
can get them from there. JPEG is preferred, cause it does not consume much
space.

What kind of format is this animation in now ? Is it FLI or FLC ?
We have the possibility to convert FLI or FLC to MPEG.
Also if you only have single pics, we can convert them to MPEG too.

In this moment MPEG encoding runs via single TGA pics, so we have to
convert your animation back to single TARGA pics and then convert them to
MPEG via the Xing encoder.

So I hope this will convince you to contact me and that you would like to
include your animation on our CD-Rom.
If you still have any questions, please send me email.

What kind of animations would make a chance? Any just-playing-around-in-my-
animations-software kind of animations or do they have to be state-of-the-art
animations looking like professionals have produced it.
Do they have to have a specific length (e.g. >30 frames) ??

No, we just take any animation which is not just a few flying dots on the
screen... We prefer funny animations, but also like technical
demonstrations or any raytracing art or simular animations done with
Animator or 3DStudio. It has not to be state of the art ! We just want to
have a Demo CD-Rom which shows what can be done with MPEG.

When are you planning to release the cd (how much time is left).

We plan to release this CD-Rom at May of this year 93 so your work should
be finaly done soon, latest submission will be around 15th of April 93.


Best regards, Stefan Hartmann, c/o Gatz & Hartmann
email to:  [e-mail]


===========================================================================
 IV | MPEG-RELATED HARDWARE
===========================

The following is excerpted from:

    VIDEO COMPRESSION OPTIONS, IEEE CICC 6-May-92

    by John J. Bloomer, [e-mail],
       Fathy F. Yassa,
       Aiman A. Abdel-Malek

[The following telephone-numbers surely are US-numbers]


-------------------------------------------------------------------------------
 IV.1 | Pipelined Processors, Building Blocks (Chip Sets)
---------------------------------------------------------

STI3220      = SGS-Thompson motion estimator (H.261, MPEG).
               602-867-6279
             - 8-bit input pixels, 4-bit H and V vectors out
             - adjustable block size matcher (8x8, 8x16, 16x16)
             - +7/-8 search window
             - 5V, 2W at 18MHz (max), 68 pin PLCC

L647*0 and
L6471*       = LSI Logic H.621/MPEG pieces.
               408-433-8000
             - L64720 motion estimator, 30/40MHz, 8x8, 16x16 blocks,
               32x32 or 16x16

TMC2311      = TRW CMOS Fast Cosine Transform Processor.
             - 12 Bits, 15 M pixels/s
             - complies with the CCITT SGXV (e.g. JPEG, H.261 and MPEG)
             - includes an adder-subtractor for linear predictive coding

HGCT         = Ricoh CRC, Generalized Chen Transform demonstration chip.
               408-432-8800.
             - 2D JPEG/MPEG/H.261 compatible DCT
             - includes quantization
             - 30MHz, 15K gates
             - licensing possible

Vision Proc. = Integrated Information Technology Inc.
               408-727-1885
             - generic DCT, motion compensated & entropy coding codec
             - microcode for still- and motion-video compression (JPEG,
               H.261 and MPEG1)
             - 1 micron CMOS, 20 MHz and 33 MHz, PGA and 84-pin QFP
             - JPEG only and JPEG/H/261/MPEG versions available, H.261 at
                   30 f/s.
             - used by Compression Labs, Inc. CDV teleconferencing system
             - rumored to be the heart of the AT&T picture phone

AVP1000      = AT&T JPEG, MPEG and H.261 codec chipset.
               800-372-2447
             - 1400D decoder, 1400C system controller
             - 1300E H.261 (CIF, QCIF, CIF240) at 30 f/s, I-frame only MPEG.
             - 1400E is superset of 1300E, motion with 1/2 pixel resolution over
               +/- 32 pixels
             - YCbCr video or digital input, on-board rate FIFOs, external
               RAM required
             - 0.75 micron, 50 MHz CMOS

82750PB,
82750DB      = Intel DVI  pixel and display YUV color space processors.
             - proprietary machine code employed for compression
             - usable for other algorithms (e.g., JPEG, H.261 or MPEG1 at
               reduced data rates)


-------------------------------------------------------------------------------
 IV.2 | Pipelined Processors, Monolithic, Fixed Lossy
-----------------------------------------------------

CL950        = C-Cube/JVC implementation of the MPEG-JVC or extended mode
               MPEG-I announced.  6-9 Mb/sec.

CL450        = Announced June 1992.  Scaled down version of CL950, with
               3Mb/sec limit. Only MPEG-I decoding.


-------------------------------------------------------------------------------
 IV.3 | Codecs Chips Under Development
--------------------------------------

MPEG1 codec chips due from - TI, Brooktree, Cypress Semiconductor, Motorola
(successor to the DSP96002 Multimedia Engine), Xing Technology/Analog
Devices, Sony and C-Cube

Windbond Electronics Corp. is developing a DSP chip for CD-I, MPEG and JPEG


-------------------------------------------------------------------------------
 IV.4 | DSP Chip Based JPEG/MPEG Solutions
------------------------------------------

Spirit-40    = Sonitech International Inc. ISA card.
               617-235-6824
             - two TMS320C40 DSPs for 80 MFLOPS
             - connect 16 boards in a hypercube for up to 1280 MFLOPS
             - JPEG, MPEG-1 audio and other voice coding applications

HardPak      = CERAM Inc., ISA and EISA file compression board.
               719-540-8500
             - 3.4 x 1.8 inch footprint (notebook, laptops)
             - 32KB on-board write-thru file compression cache
             - CERAM also has an SBus compressive swap-space accelerator for
               Suns

macDSP       = Spectral Innovations, AT&T DSPC32-based accelerator.
               408-727-1314
             - JPEG functions available
             - 30 MFLOPS on the NuBus


===========================================================================
 V.1 | MAILBOX-ACCESS
=====================

This is the phone number of Xing Technologies' BBS:

  805-473-2680 (2400b) (USA)

Bryan Woodworth <[e-mail]> wrote:

Would you also please add, that the Xing BBS now supports v.32bis and HST !
I am not sure on HST, but I am sure it supports v.32bis.  However, I have a
v.32bis modem, and could only connect at 9600. I think they do not have the
modem configured properly.

[ Well, Xing's software is dominating the MPEG-market, so what's about a ]
[ Internet-Connection ?                                                  ]

---------------------------------------------------------------------------
 V.2 |
------

These are the phone numbers of Gatz & Hartmann's
7 line support BBS:

  ++49 30- 462 63 41 (v32bis)
  ++49 30- 462 64 35 (v32bis)
  ++49 30- 462 65 38 (v32bis)
  ++49 30- 462 60 22 (v32 + PEP)
  ++49 30- 462 61 37 (v32)
  ++49 30- 462 62 37 (v32)
  ++49 30- 461 86 50 (v22bis + HST)

This is the professional Zelator-ACCESS-BBS system with Internet access.
There will be several new MPEG clips and updates of the GENOA 7900 SVGA
board drivers, 24 bit ET4000 programing infos,etc... Check it out ! You
will enjoy it.

Just log in with:

  guh

That means: Gatz und Hartmann.


===========================================================================
 VI.1 |  FTP-ACCESS (PD)
========================

There is an MPEG archive site at:

  phoenix.oulu.fi (130.231.240.17) in the directory
  /pub/mpeg

Here is the current list from /pub/mpeg:

-rw-r--r--  471502 Sep 13 17:36 MPEGXING.LZH
-rw-r--r--    1192 Oct  2 21:48 TUTTIF3D.DOC
-rw-r--r--  502473 Jul 23 21:53 birdisba.mpg
-rw-r--r--     696 Jul 23 22:25 birdisba.txt
-rw-r--r--  233981 Jul  7  1992 joel.lzh
-rw-r--r--    1137 Jul  7  1992 joel.txt
-rw-r--r--   34283 Jul  7  1992 lha.exe
-rw-r--r--     278 Jul  7  1992 lha.txt
-rw-r--r--  292665 Jun 25  1992 moglie.mpg
-rw-r--r--     439 Jun 25  1992 moglie.txt
-rw-r--r--  244095 Sep 18 12:42 mpegplay-020792.lha
-rw-r--r--  368955 Sep 23 00:30 mpegplay.zoo
-rw-r--r--  721801 Jun  3  1992 mpgmovie.lzh
-rw-r--r--     368 Jun  3  1992 mpgmovie.txt
-rw-r--r--  978660 Sep 13 17:35 raiders.mpg
-rw-r--r--  250937 Jul  4  1992 rom.mpg
-rw-r--r--     951 Jul  4  1992 rom.txt
-rw-r--r--  534405 Jul  3  1992 sukhoi.mpg
-rw-r--r--     342 Jul  3  1992 sukhoi.txt
-rw-r--r--  414427 Oct  2 21:45 tuttif3d.lzh

Please contact this ftp-site for files before e-mailing to me !!!


---------------------------------------------------------------------------
 VI.3 |
-------

There is an MPEG archive site at:

  toe.cs.berkeley.edu (128.32.149.117) in the directory
  /pub/multimedia/mpeg


---------------------------------------------------------------------------
 VI.3 |
-------

Gatz & Hartman BBS is now reachable via ftp, between 18.00 - 6.00 german
time. Login as 'gast', then look for IBM-Files under File-Sector 14 :
IBM_g_und_h

  zelator.in-berlin.de (192.109.42.11)


---------------------------------------------------------------------------
 VI.4 |
-------

Bryan Woodworth <[e-mail]> invites you to the ftp-server:

    ftp.rahul.net (192.160.13.1) in /pub/bryanw/pc/animation/mpeg

Login as "anonymous," any time of the day or night.

[ Several MPEG-Information is located in the directory /pub/bryanw   ]
[ Bryan was the first one, that downloaded the brand new mpeg-player ]
[ from Xing's BBS and posted it to a.b.p.u, thnx to Bryan !          ]

He wrote:

If the people have problems connecting, they should send a capture of the
session to "[e-mail]," so that the problem can be corrected.


---------------------------------------------------------------------------
 VI.5 |
-------

Good ftp-hosts to look for MPEG-related software are:

  ftp.cica.indiana.edu (129.79.20.84)
  ftp.germany.eu.net (192.76.144.75)
  ftp.cs.tu-berlin.de (130.149.17.7)
  mucket.vast.unsw.edu.au
  nic.funet.fi (128.214.6.100)
  pinus.slu.se (130.238.98.11)
  wuarchive.wustl.edu (128.252.135.4)


===========================================================================
 VII |  MAIL-ORDER
==================

                   GET THE "TRAIL-PACK" !!!
                  ==========================

You can purchase a complete archive named the "Trail-Pack" including the
FAQ and all named programs, source-code, movies and information-files.

This archive includes (in addition to the ftp-access) all versions of the
programs and source-code, additional movies (including the audio-wav-files)
and lots of additional informations.

It will contain at least

         4,094,228 AUDIO
         6,983,207 BIGMOVIE
           836,301 CODEC
           913,193 DEMO
         1,501,498 DOC
           270,728 FAQ
        17,498,558 MOVIES
         1,430,828 UTIL
      ======================
Total = > 33 MB compressed = > 50 MB uncompressed MPEG !

To obtain the "Trail-Pack" send a envelope, with the big-written
key-word "Trail-Pack" on it to:

      PHADE SOFTWARE
      Inh. Frank Gadegast

      Leibnizstr. 30
      1000 Berlin 12

      G E R M A N Y

and include in it:

 o  30 DM (thirty German Mark), to pay the time I spend on copying disks
    and going to the post (money, that will be over, will be used to
    prepare the next version of this FAQ). Please do NOT included ANY
    coins.

 o  25 HD-floppy-disk (3-1/2" or 5-1/2"), allready formatted (MSDOS).
    If this FAQ is older than about 3 month (please look at the date
    at the top) included a few MORE floppy-disk, because the archive is
    growing day by day.

 o  a hard-cover-envelope (big and strong enough to carry the 25 floppy-
    disks, written with YOUR correct adress.

 o  enough money (at least 15 DM from other countries, 10 DM from inside
    Germany) to pay the postage of the "Trail-Pack" (the postage of your
    package to me, should be nearly the same, compared to, what I have to
    spend, to send the "Trail-Pack"; so, if you are living in Australia,
    send MORE money; the last package back to Taiwan cost about DM 24 !).
    Again, please do NOT include ANY coins.

Try to send me a ENVELOPE, not a packet. It should go through the letter
box.


NOTE: There is no guarantee, how, and when you will get the "Trail-Pack"
      back. I'll do my best to prepare the packages as quick as possible.
      But I can't guarantee for the post ;o)

NOTE: Please do not send any schecks, or try to pay via credit-cards.

NOTE: Requests, that are NOT complete will be send back, using the
      included money. Is no money included, nothing will be send back
      and my archive will thank you for your floppy-disk-gift !!!
      Is there no money neither floppy-disks, your envelope will go
      to the bin.

NOTE: This is NOT a commercial offer, it's a service for those, that
      don't have internet-access !!!


===========================================================================
 VIII |  RETRIEVED MAIL
=======================

From: [e-mail] (KETAN DASHARATH PATEL)
Subject: Re: Xing's SW, Really MPEG Compression?
Date: Thu Nov 19 13:20:35 1992

Unfortunately, it is true. XING data is NOT true MPEG and in
fact does a lot of dubious things with its Inverse DCT.

XING data is simply a sequence of I-Frames (i.e. no interframe
compression is done, no motion vectors, nothing).
This amounts to little more than a sequence of JPEG type images.

Ketan Patel
[e-mail]

---------------------------------------------------------------------------

From: Mark Adler <[e-mail]>
Date: Mon, 30 Nov 92 11:36:19 -0800
Subject: Re: MPEG - FAQ

There was a correction in the trade journal where I saw the MPEG II
extension mentioned.  They are now saying that, yes, they did agree to
extend MPEG II's target bit rate to meet the needs of HDTV applications,
but that there is no real upper limit to the MPEG II compressed rate, and
that they expect no delays in the completion of the final spec, scheduled
for March 1993.

---------------------------------------------------------------------------

Those of you working on MPEG encoding might be interested in the
accelerator board we have built for computing motion vectors and DCTs on
Sun workstations.  Following is a description.  For ordering information,
contact me at:

     Space Computer Corporation
     2800 Olympic Blvd.
     Suite 104
     Santa Monica, CA 90404
     Telephone: [phone]
     FAX: [phone]
     Internet: [e-mail]

Bill Kendall

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

     VCA-1 VIDEO COMPRESSION ACCELERATOR FOR SUN WORKSTATIONS

      Special-Purpose Hardware for Motion Estimation and DCTs

------------------------------------------------------------------
FEATURES:

Performs 8x8 DCTs in 21 microsec after first DCT at 52 microsec.*
Performs 32x32 cross search for 16x16 block in 239 microsec.*
Mounts in a single SBus slot.
Included software allows user-transparent access.
Price: $2,900 (subject to change without notice).

*Stated times are for a 25-MHz SBus.
------------------------------------------------------------------

The VCA-1 Video Compression Accelerator is a one-slot SBus add-in
board for Sun workstations which reduces the run time for MPEG encoding
up to an order of magnitude.  It is intended to aid both researchers
and designers of MPEG-based digital video products.

The VCA-1 accelerator can be utilized in any SBus-based workstation.
The VCA-1 provides high-speed hardware execution of the two most
computationally-intensive MPEG compression operations: motion estimation
and discrete cosine transform (DCT) computation.  These operations
are performed by two special-purpose VLSI devices manufactured by
LSI Logic Corporation.

The first device is the L64720 Motion Estimation Processor, which
contains 64 high-performance, special-purpose processors on a single
chip.  This device implements an exhaustive search algorithm in which
a user-selectable data block in the current video frame is offset
and compared with a window in a reference image.  The device computes
the position of the best match, the minimum error, and the zero-offset
error.  The error computed is the sum of absolute differences.  The
data block is user-selectable for either 16x16 or 8x8 pixels, and the
size of the search window is adjustable by setting software parameters.

The second device is the L64730 Discrete Cosine Transform Processor,
which contains eight special-purpose processors.  This device computes
both the forward and inverse DCT over 8x8 blocks, delivering 12-bit
output data for the forward transform and 8 or 9-bit output data for
the inverse transform.  It also can be used for loop filtering and
inter-frame prediction subtraction.

In addition to the above devices, the VCA-1 contains 3 MBytes of
DRAM and three gate arrays.  The VCA-1 is fully compliant with SBus
standards, and requires only a single workstation slot.  If all of
the SBus slots are already filled, the VCA-1 can be installed in an
inexpensive expansion box sold by Sun Microsystems.

Software provided with the VCA-1 accelerator consists of routines
which completely control the board's various functions.  This software,
which runs under Sun OS, can easily be integrated with existing MPEG
encoding software to perform motion estimation and DCT/IDCT computations.
Full source code is provided.

---------------------------------------------------------------------------

From: Bryan Woodworth <[e-mail]>
Subject: Re: MPEGFAQ.INFO -- Thanks, Frank.
Date: Wed, 24 Feb 93 12:12:59 PST

In other news, Xing is working hard on a DOS MPEG player with WAV file
support.  It should be out (it had better be..) within a year, most likely
3-4 months.  It will work via VESA autodetect; I suspect it will use VESA's
640x480 mode.

---------------------------------------------------------------------------

From: [e-mail] (Todd Brunhoff)
Subject: New product for MPEG, JPEG compression under UNIX and X
Date: 16 Feb 93 01:08:43 GMT

North Valley Research announced today immediate availability of a family of
products for working with video and other time-based media in a UNIX
environment.  These products enable the end user to take video and audio all the
way from video camera or tape to an MPEG sequence that can be played back in
real-time on most Sun SPARCstations, without hardware assistance.

For more info:
  - reply to me personally (don't post a followup)
  - get the announcement via anonymous ftp to nvr.com (192.82.231.50) in /pub/NVR
  - check comp.windows.x.announce, comp.newprod, comp.sys.sun.announce
---------------
internet: [e-mail]                                             c--Q Q
US:       Todd Brunhoff; North Valley Research;                         `
          15262 NW Greenbriar Pkwy; Beaverton, OR  97006                -
Phone:    [phone]
Fax:      [phone]

---------------------------------------------------------------------------

From: [e-mail] (Rainer Menes)
Subject: MPEG for Mac Info
Date: Fri, 19 Mar 1993 12:15:54 GMT

Dear MPEG for the Mac users,

Today I think more than 200-300 user are out there using the player. I am
not getting any big response to my software. So I think it works ok. Does it
realy ??? I must read in Comp.sys.mac.digest "the slow and buggy player".
Slow is ok but buggy? Are there any bugs which I don't know? If there any
besides it is slow or the file type is incorrect, or the mpeg file are
altered during transfer to the Mac (lf->cr), then please mail me, and
ftp the buggy mpeg file to suniams1.statistik.tu-muenchen.de (131.159.64.1)
and put it in the dir /incoming/buggy_mpeg. Thanks!! I could only fix these
problem when I know them.

Now I have collected some tools which might be usefull for most of the users:

1. bunchtyper 1.01 an util which takes care of the TYPE and CREATOR of a
file or a collections of files. This has shown to be the bigest problem at
all with my mpeg player.

2. filetyper 3.2 the same as bunchtyper (take both and decide what do you like
best).

3. picttomovie 1.0 converts the picts created by the converter to an
quicktime movie. Some users have quicktime but not convert to movie. This
utils should closes this gap.

4. qtbatchcompress1.0 a util to convert the picts to jpegs-pict with
quicktime. This util might be usefull to reduce the storage use by tones of
pict-files from a converted mpeg-movie. ( Trying differnt codecs or what
ever).

You will find the tools in the dir /pub/mac/utils on my ftp-site.
(suniams1.statistik.tu-muenchen.de 131.159.64.1 or all other major archive
sites have these tools too but not in the same dir you have to look for it)

Now to the problem with the slowness of the software. I first got aware of
that problem two days ago when I meet a friend of mine who has a MacII si.
Some files are running only with 0.5 - 0.8 frames a second. Sorry I did't
know. That the differnce between the quadra 700 and the si is up to 10 times
in speed differnce I have become aware of, only two days ago. My Quadra plays
the same movie with 5.3 - 7 frames a second. I don't what to know what
frame rate a Clasic II or a LC may reach.

Now how does the player compare to other computers running the same
software (sourcecode is the  Berkley MPEG-Player for X-Window):

486-33MHz 256KByte  extern-cache with Linux 0.99.7 gcc 2.33 -O2 -m486 7.3
frames/sec.

Quadra 700 0KByte extern-cache with System 7.1 gcc1.37r7 -O2  5.2
frames/sec.

Sparc Station SLC ?? with sunos 4.2 gcc 2.33 -O2 6.7 frames/sec.

SGI Iris Indigo R3000 with Unix 4.5 cc -O 15.0 frames/sec.

all for the same movie flight.mpg but with differnt framebuffer depth. This
may slightly infuence the results.

The most interresting value is the differnce between the 486 and the 68040.
First both processor have a very close architecture (cache, some
RISC-features and so on). Even the efficents how many cycles per instruction
is very close. If I remember correct the 68040 and 486 have 1.3 - 1.2 cycles
per instruction in the mean. Now you see it would be very surprising if the
68040 25MHz is faster than a 33MHz 486 which has an external cache. The
difference in frame rate is nearly simular to the difference in clock speed.
Another point is the gcc compiler has made a drastic improfment between
version 1.37 and 2.33 which was used on the 486 with Linux (a freeware Unix
clone). There are no optimisation done for the 68040 at all but as you see
the Linux version is optimised for the 486. I would expect to get 10- 20%
speed improfment with gcc 2.33 for the 68040. Beside I made a new version
which has been speeded up alittle 15-20% and the gcc 2.33 compiler should be
out in a month or so. This makes me optimistic than my Quadra will play back
a mpeg film as fast as a normal 486-33Mhz in the near future.

The new version should be out next week. It will support any resolution of
a MPEG-film. As said befor it is alittle faster and the converter will use the
filename of the orginal mpeg film as name and not test.xxxx. Only small
improfments but Rom wasn't build in a day.

Hope you have fun with my player, and you report bugs to me

Til next week,

Rainer

email: [e-mail]

---------------------------------------------------------------------------

From: [e-mail] (Bryan Woodworth)
Date: Mon, 15 Feb 1993 21:47:18 GMT

In <[e-mail]>
[e-mail] (Marcus Ahlfors) writes:

>Wanted a MPG player for Amiga that fully works. Not the MPEG_Viewer packet
>what needed some strager library that noone had.. A packet that fully works
>thanks.

I know nothing about Amiga (except they are very nice for multimedia) but
someone uploaded some mpeg players to my site, you may want to try..

ftp.rahul.net, 192.160.13.1, /pub/bryanw/amiga/...

---------------------------------------------------------------------------

From: [e-mail] (Michael van Elst)
Subject: Re: MPG Viewer for Amiga Wanted!!
Date: 17 Feb 93 01:23:05 GMT

I once uploaded a version mpeg_play1.2 to amiga.physik.unizh.ch under the
name mp.lha.

I just got mpeg_play2.0 working but don't want to upload until I've finished
HAM support.

If you can live with EHB-only please contact me.

---------------------------------------------------------------------------

From: [e-mail] (Michael van Elst)
Subject: Re: MPG Viewer for Amiga Wanted!!
Date: Wed, 17 Feb 1993 17:07:44 GMT

If nothing goes wrong I'll upload my player (which supports Amiga EHB
and HAM6 graphics) next week to one of the aminet sites.

---------------------------------------------------------------------------

Date: Tue, 2 Mar 93 17:25:06 -0800
From: [e-mail] (Larry Rowe)
Message-Id: <[e-mail]>
Subject: Re: REPOST: THE MPEG-FAQ - Version 1.1 [1/2]

The ones [decoder] I can disclose are from Sony and one from
North Valley Research (contact Todd Brunhoff - [e-mail]).

[ How about a free encoder running on Unix ? Where yous not working on it ? ]

We are working on a freely distributed mpeg video encoder.  But, it still has
a ways to go.  We're trying to produce a workable system quickly without too
much focus on speed or ultimate coding efficiency.  It will be available
sometime within the next 3-4 months.

---------------------------------------------------------------------------

From: [e-mail] (Jerry Pierce)
Subject: Re: MPEG on silicon
Date: Wed, 17 Feb 1993 16:57:23 GMT

E: REAL time MPEG encoder

The GCT MPEG encoder announced in EETimes was designed by Nippon
Steel and my company, Eidesign Technologies. (The relationship
between ASCII, GCT, Nippon Steel, Hitachi, JVC is a long and
complicated one--suffice to say the units have been designed and a
few have been sold and DELIVERED.)

We have one operating at Eidesign's headquarters in Mountain View,
CA. The design is a three board set. The algorithm is Nippon Steel
proprietary and is a FULL MPEG solution--I, P, and B frames. The
resultant quality, in real time, is very good. We have compared
favorably to C-Cube's non-real time algorithm. (I don't know how
we will compare to their chip set.) We are adding a MPEG level 2
audio solution to the system (it is currently being implemented
and not yet working--March?).

The encoder box accepts CCIR 601 input (D1) and has a SCSI port to
connect to a workstation for assembling the system layer. The box
creates a complete MPEG video stream in REAL time (a delay of
about 1 second from the time the video is presented to the box and
the time it appears decoded by the decoder--H.261 it's not!) We
are now using a Thomson MPEG video decoder to monitor the results.

We (the design team is split between Eidesign and Nippon Steel)
decided to build a flexible MPEG encoder using off-the-shelf chips
and use it as soon as possible. This includes DSP's, FPGA's,
motion estimation chips, etc. It IS flexible and we COULD
implement other algorithms, however we are not using it as a
research tool, but as a solution to acquisition of MPEG material
for a variety of platforms. I suppose it could be used as a
research tool.

I don't know where/how they came up with the $250K price, but we
are discussing the system with interested parties for specific
configurations to meet their needs. A FULL up system with audio,
video, and other design tools may cost this much.

Jerry Pierce
Eidesign Technologies
1923 Landings Drive
Mountain View, CA 94043
415-903-9211 x15
[e-mail] or [e-mail]


===========================================================================
 IX | ADDITIONAL INFORMATION
============================

From: [e-mail] (rakeshkumar gandhi )
Date: Tue, 24 Nov 92 13:14:03 -0500
Subject: IEEE

There is MPEG Hardware review in IEEE computer graphics and application
magazine.

---------------------------------------------------------------------------

Only for Germans:

Ihr koennt den MPEG-draft-I beim Beuth Verlag bekommem.


===========================================================================
 X | WHERE TO FIND MORE INFOS
=============================

Well, first you can check the related news-groups:

  comp.graphics, comp.graphics.animation, comp.compression, comp.multimedia,
  comp.sys.amiga.multimedia, comp.mail.multi-media,
  alt.binaries.pictures.utilities

The first part of this FAQ about MPEG came from Mark Adler, published in
in FAQ for the newsgroup 'comp.compression'.

---------------------------------------------------------------------------

Then you can ask 'archie' to find all NEW mpeg-releated software
by sending the following mail (with no title):

  prog mpeg mpg
  quit

to one of the following archie-mail-servers:

  [e-mail]
  [e-mail]
  [e-mail]
  [e-mail]
  [e-mail]
  [e-mail]
  [e-mail]

---------------------------------------------------------------------------

Then you could look for a newer version of the first part of this FAQ via
ftp at:

    garbo.uwasa.fi (128.214.87.1), in /pub/doc-net

    The current version is named FAQC9301.ZIP

---------------------------------------------------------------------------

Then read this (oh Bryan, you're gas ...):

From: [e-mail] (Bryan Woodworth)
Subject: MPEG VIEWERS FOR MSWIN/XWIN/VMS/AMIGA/MACINTOSH AVAILABLE!
Date: Tue, 30 Mar 1993 21:26:37 GMT

Last Updated:  30.3.93   12:32:34

     MPEG VIEWERS FOR VARIOUS PLATFORMS:  OBTAINMENT INFORMATION


a.  Microsoft Windows MPEG viewer info
     i. Xing's MPEG Player << Obtainable via email
    ii. An exciting port of the Berkeley MPEG player << Also via ftpmail
b.  Unix X-Windows viewer info << Also via ftpmail
c.  Macintosh viewer info << Also via ftpmail
d.  VMS viewer info << Also via ftpmail
e.  Amiga viewer info << Obtainable via many means besides ftp
f.  Reminders when obtaining files
g.  Information on obtaining files via email
h.  Where to obtain MPEG files
i.  Conclusion

a.  Windows MPEG viewer

There are two Windows MPEG viewers.  One is released by Xing
Technology as freeware; the other is a Shareware viewer provided by
a private party which is a port of the Berkeley v2.0 freeware MPEG
player.

First, I will detail Xing's MPEG player for Windows.

i.  Xing's freeware MPEG player for Windows, version 2.0.

I am not certain if it works with Windows 3.0, but who cares,
right?  Everyone uses Windows 3.1!  And I know it works with 3.1.
:-)   Once you have Windows all you need is any SVGA card.  The
Windows MPEG player works with ANY SVGA card, provided it has a driver for
640x480x256 colors in Windows 3.x.

Read HELP.INSTALLING.MPEG.FOR.WINDOWS for more information (see below).

You may obtain the viewer via toe, simtel20 mirrors via ftp
(see below) or email (see SIMTEL20.EMAIL at ftp.rahul.net [192.160.13.1] in
/pub/bryanw/information, or next posting if you are reading this in a
Usenet newsgroup).  Herewith is the ftp info:

You can get the viewer from:

Site:  toe.cs.berkeley.edu [128.32.149.117]
Dir : /pub/multimedia/mpeg/Windows3.x
File: mpegexe.zip + attendant dll (such as mddlati.zip)

The viewer is also available from all simtel20 mirrors.  One
specific site I can remember is as follows:

Site:  oak.oakland.edu
Dir :  /pub/msdos/windows3
File:  mpegexe.zip, + attendant dll (such as mdllati.zip)

Here is a listing of all simtel20 mirror sites, from a recent posting by
[e-mail] (Keith Petersen) to comp.archives.msdos.announce.
Please remember to use the ftp site closest to you geographically:

SIMTEL20 files are also available by anonymous ftp from mirror sites
OAK.Oakland.Edu (141.210.10.117), wuarchive.wustl.edu (128.252.135.4),
ftp.uu.net (137.39.1.9), nic.funet.fi (128.214.6.100), src.doc.ic.ac.uk
(146.169.3.7), nic.switch.ch (130.59.1.40), archie.au (139.130.4.6),
nctuccca.edu.tw (140.111.3.21)....

Please read HELP.INSTALLING.MPEG.FOR.WINDOWS, located on
ftp.rahul.net, in /pub/bryanw/information, whence you obtained this file.
It gives tips on installation and answers many questions.  IP address for
those whose servers don't accept names:  ftp.rahul.net, 192.160.13.1

I spent a great amount of time compiling it, so I will hold a very
low opinion of anyone who asks me for help without having read the
HELP.INSTALLING.MPEG.FOR.WINDOWS file first.

However, if there is a question you have which is not answered by the
HELP.* file then please do send me email so I can add it to the HELP* file.

ii.  Berkeley MPEG player Port:  Supports 15bit-24bit displays!

The following is a verbatim transcription of an email message I
received about this other player.  It is worth a look.

--- begin ---
~Date: Mon, 29 Mar 93 10:00:08 MST
~From: [e-mail] (Russell Webb)
To: [e-mail]
~Subject: Re: FTP SITE AVAILABLE FOR MSDOS GRAPHICS/UNIX UTILITIES
Organization: Nyx, Public Access Unix @ U. of Denver Math/CS dept.

I'm sorry to have seen your archive turned into a mini-archive, but
your effort is still apreciated.  Regarding the above information
you currently provide: there is another MPEG player that you may
have overlooked so far.  Check toe.cs.berkeley.edu under
/pub/multimedia/mpeg/ mpegnt.zip for a new (~March 15) Windows NT
MPEG viewer, a port of the Berkeley Unix v2.0 player.  It *will*
run under Windows 3.1, but the Win32s subsystem is required.  The
Win32s package is bundled with the Spice32/ Nutmeg32 circuit design
package that can be found on the ftp.cica.indiana.edu site (under
/pub/pc/win3/nt/spice4?.zip, I believe).  [ED. NOTE:
ftp.cica.indiana.edu's IP address is 129.79.20.84]

Why go through all the hassle of grabbing this if you don't have
Windows NT?  Because the player supports 15-bit and 24-bit color.
It looks quite nice in comparison to the Xing player or an 8-bit
X-Windows display. It's a bit slow (on my 486/50, 16M ram) and easy
to crash--but still the only free way I know of to get a
hi-color/truecolor MPEG display on a PC (short of having a 15- or
24-bit X terminal on Intel hardware).

I hope this information is a bit useful.  Thanks again for your
mini-archive and information postings.

-Russell Webb
[e-mail]
--- end ---

b. Unix X-Windows MPEG viewer

There is also an MPEG player for Unix X-windows platforms, which
supposedly works fairly well.  When I had Linux running on my
386/25 with X Windows, it worked fairly slowly, but this WAS an IBM
80386, no math compressor, and "only" 8 megs of ram.  :-)   Perhaps
you will fare better.

The term "X-Windows" encapsulates any computer which can run X-Windows
software.  These include Unix machines such as Suns, and other machines
made by Hewlett Packard, etc.  IBM computers can also emulate X-Windows
using the Linux or 386BSD Operating Systems along with a port of the X
windows environment.

For Linux information read comp.os.linux; 386BSD,
see the following newsgroups for information on 386BSD:

comp.os.386bsd.announce
comp.os.386bsd.apps
comp.os.386bsd.bugs
comp.os.386bsd.questions

Site info:

Site:  toe.cs.berkeley.edu [128.32.149.117]
Dir :  /pub/multimedia/mpeg
File:  mpeg-*

c.  Macintosh MPEG viewer

The Berkeley MPEG player seems to have provided the impetus for MPEG
viewers on every system!  The Berkeley MPEG viewer has now been ported to
the Macintosh platform as well.

Site info:

Site:  toe.cs.berkeley.edu [128.32.149.117]
Dir :  /pub/multimedia/mpeg/mac
File:  mpeg_mac_*

d.  VMS MPEG viewer

Yet another port of the Berkeley MPEG player.  Contains specific
source for VMS machines which will aid in compilation.

Site:  toe.cs.berkeley.edu [128.32.149.117]
Dir :  /pub/multimedia/mpeg/vms
File:  Browse entire subdir, snag what you need

e.  Amiga MPEG viewers

There are many MPEG viewers for the Amiga.  They are all available via
ftp from amiga.physik.unizh.ch [130.60.80.80] or one of its many mirrors.
It is highly recommended you use the mirror geographically closest to you
to reduce network bandwidth and increase transfer time.  Here is a listing
of all mirrors (including amiga.physik itself):

Switzerland  amiga.physik.unizh.ch  130.60.80.80    pub/aminet/
Switzerland  litamiga.epfl.ch       128.178.151.32  pub/aminet/
Scandinavia  ftp.luth.se            130.240.16.3    pub/aminet/
Germany      ftp.uni-kl.de          131.246.9.95    pub/aminet/
Germany      ftp.uni-erlangen.de    131.188.1.43    pub/aminet/
Germany      ftp.cs.tu-berlin.de    130.149.17.7    pub/aminet/
Germany      ftp.th-darmstadt.de    130.83.55.75    pub/aminet/
Germany      ftp.uni-paderborn.de   131.234.2.32    pub/aminet/
USA          ftp.wustl.edu          128.252.135.4   pub/aminet/
USA          merlin.etsu.edu        192.43.199.20   pub/aminet/
USA          oes.orst.edu           128.193.124.2   pub/aminet/
Australia    splat.aarnet.edu.au    192.107.107.6   pub/aminet/ (*)
   (*) closed 6:30am to 4pm weekdays

Here is a listing of filenames from the LONG.Z file.  Merely cd to the
proper subdirectory (in this case, /pub/aminet/gfx/show) and get what you
need.

                                      Days
Filename             Directory   Size/Old  Description
--------------------------------------------------------------------
mp.lha               gfx/show    45K  83 MPEG player for EHB display.
                                         Needs OS2.0
mpeg2_0amiga.lha     gfx/show    50K  40 Berkeley MPEG player 2.0
mpegplay201_bin.lha  gfx/show   147K  43+MPEG player V2.01 executable
mpegplay201_src.lha  gfx/show   170K  43+MPEG player V2.01 sources
mpeg_player122.lha   gfx/show   206K 104+MPEG Player 1.22 (for all Amigas)

You can also obtain the software via aminet using other means.  From the
README:

                         OTHER AMINET ACCESSES
                         ---------------------

There are many other ways than FTP to access AmiNet:

- ADT. This is a front end for FTP that allows easy access to AmiNet.
  Get it from comm/misc/ and compile it on your UNIX box.

- FSP. AmiNet Files can be downloaded from the FSP site disun3.epfl.ch
  port 9999. Uploads are accepted and forwarded.

- NFS. The only AmiNet site that allows NFS mounting of the archives is
  wuarchive.wust.edu. FTP there and read the details in the /README.NFS

- IRC. On Internet Relay Chat, you can talk to various server robots
  like AmiBot, MerBot or Mama, to do queries and retrievals.

- Gopher. There is a gopher server for AmiNet at merlin.etsu.edu. To
  connect, use the command 'gopher merlin.etsu.edu'.

- Modem. In Germany, you can download the AmiNet files from the Incubus
  BBS, telephone number 0931 781464. The login is 'ftp', password 'ftp'.

- Usenet. A list of recent uploads is posted every week to the newsgroups
  comp.sys.amiga.misc and de.comp.sys.amiga.misc. Useful for mail servers.

- Mailserver. Sorry, no specialized e-mail server for AmiNet yet. But you
  can use [e-mail]. Send a mail with HELP in the body.

- CD-ROM. AmiNet is available on CD-ROM. Talk to [e-mail], or write
  to Walnut Creek CDROM, 1547 Palos Verdes Mall, Walnut Creek CA 94596, USA
 or phone [phone] or [phone] (FAX)


f.  Reminders..

o Login anonymously.  I.E.

FTP server Omygosh-v2.21 "It's in there!", <c> P in a Square Ltd READY.
login:  anonymous
password:  (enter your email address here, like "[e-mail]")

...then you're in.

o Use binary mode when transferring files from ftp sites.
  I.E. "type binary" or "binary" before transferring a file.

E.g.

ftp> binary
Binary mode set.
ftp> get mpegexe.zip
...

g.  Obtaining files via email

Some of the sections above cover specific information for retrieving files
via specific means, such as for the MSWIN Xing player (via simtel20's
listservers) or the amiga (via other various means outlined in that
section).  For the remaining sections, it is possible to use the ftpmail
server.

FTPMAIL, obtaining files from a server via email which does the ftping for
you, is not the best way to obtain files via ftp, but for some it is the
only way.  To get more information, send email to [e-mail]
with the word HELP in the message body.

h.  Where to obtain MPEG files

Now that you have your glorious MPEG viewer, where to obtain MPEG files to
test your program?  Luckily there are a few sites whence many MPEGs may be
obtained.

The best site for MPEGs is toe.cs.berkeley.edu [128.32.149.117], which has
a veritable slew of MPEGs available in /pub/multimedia/mpeg/movies, and a
glorious MPEG MEGA demo, crafted by Stefan Hartmann
([e-mail]) in /pub/multimedia/mpeg/genoa.

Another good site for MPEGs is phoenix.oulu.fi [130.231.240.17], but please
use this site sparingly unless you are right next door to Finland --
bandwidth is getting expensive these days!   Most files available on
phoenix are already at toe anyhow.

Have fun!

i.  Conclusion

MPEG is a fascinating animation format which is compact yet offers lots of
animation for the space!  With ports of the MPEG player available now for
most major platforms, MPEG can now be shared by almost everyone.  I hope
you will enjoy MPEG as much as I have.  Sometimes it's annoying searching
for a viewer for your platform, but it's all worthwhile in the end.

-Bryan

--
Bryan Woodworth <[e-mail]>


===========================================================================
 XI | NEWS
==========

Well, here the news from the CEBIT'93 in Hannover GERMANY:

Nothing special happened (except the new NeXT-Station with Intel-
Board, Expansion-Slot for the Pentium and Localbus running Mach486,
X11 and Windows 3.1 under the same GUI: NeXTStep) and except the
MPEG-News:

Brainstorm was showing a Beta-Version of a player for the Atari.
Its doing I-, P- and B-Frames and could perfectly play all the Xing-Movies.
(Not sure about the big ones).

More infos from: [ and read Section III.7 ]

  Raphael Lemoine          at

  Brainstorm
  19 bis
  Rue de Cotte
  75012 Paris
  FRANCE

  Vocal:   1 - 44 67 08 09
  Fax:     1 - 44 67 08 11
  Modem:   1 - 44 67 08 44
  Fidonet: 2:320/100.4

---------------------------------------------------------------------------

Leadtek was showing there DOS-full-screen-MPEG-player. They double
the pixel in a tricky way, so they get 640x400 and the quality is
really good. They told me, the player (with lots of additional
software) is to buy for about $900. The contact address is:

  Mr. Terry Yeu          at

  Leadtek Research Inc.
  Computer Graphics, Multimedia Design & Manufacture

  5F, NO. 4, Alley 11, Lane 327, Sec. 2, Chung-Shan Rd.,
  Chung-Ho, Taipei Shien,
  TAIWAN R.O.C.

  Tel: [phone] Ext 113
  Fax: [phone]

---------------------------------------------------------------------------

Sun has a new version of there 'Multimedia Solutions for Workgroup'
out. And (but this is not official), they will support MPEG, but this
was not to be seen.

---------------------------------------------------------------------------

[ A new coding sheme for video arrived. It has NOTHING to do with MPEG,  ]
[ but it produces a fabulous frame-rate, the compression is much better, ]
[ but therefore the quality is less than MPEG. Here the news:            ]

Date: Mon, 15 Mar 93 13:58:29 EST
From: [e-mail] (Ruben Gonzalez)
Subject: Re: riscdemo- more infos please ...


> > We aim to solve these problems with our new technique. We get up to 40:1
> > compression (and rising) and frame rates of up to 100+ frames per second
> > on a standard 486 IBM PC with a standard VGA card. No fancy anything
> > is required.

We've actually improved the compression quite a nit lately.

> How does it work ?

Well basically it does a bit of vector quantisation and we try to decompose
the images into graphic primitives.

> 100 frames of which size (how many colors) ?

120 fps at 176x144, we get about 35 fps at 352x288. All in 256 colours
but there is no contouring artifacts as with *all* the other schemes that exist.

> How do you compare it to MPEG ?

Well they get a bit more compression, but their quality is quite a bit lower.
Also the software MPEG version is not full MPEG just a portion of it. MPEG
needs to do a DCT transform in reall time which is really time consuming,
MPEG also need to do real time colour quantisation and we don't need to do
any of this in fact our decoder is only about 50 lines long :^)

Our version is fully extensible into hypermedia and full object level
interaction, whereas the MPEG stuff is restricted to only multimedia with
no intra frame- object level interaction.

> but how does it really work, whats the algorithm to reach such a compression
> without a DCT (f.e.) ?
>

Hey, a DCT does not give any compression either, it's the quantisation of and
succeeding variable length (Huffman) coding of the transform coefficents
that give rise to the compression. Of course the motion compensation also
helps.

Well as I said its a combination of adaptive vector quantisation and graphic
primitive decomposition. We also make heavy use of human visual system's
characteristics. To explain the entire algorithm would take up many pages
since it is completely original and there is nothing I could sort of
compare it too. Plus I'm not allowed to reveal certain parts of the algorithm.

For further information please refer to :

R.Gonzalez and A.Ginige, "Towards Integrating Digtial Video for Multimedia",
Australian Broadband Switching and Services Symposium'92, Melbourne Australia.

A.Ginige, A.Serevanute, R.Gonzalez et.al. "An Experimental Multimedia System"
Computer Communication Review, Vol.22, No.3, July 1992.

There are also other published papers on the algorithm floating around.

> Could you maybe send me some of your compressed video's ? Or tell me, where
> to ftp from ? Are there any documents, describing the technic ?

You should be able to ftp some videos from  : ftp.uts.edu.au under the
directory /pub/RISCvideo

I would compress some more videos but I'm busy working on the new improved
version as well as porting the algorithm to MSWindows and other platforms
and there is only one of me :(.


Ruben Gonzalez.
School of Electrical Engineering,
University of Technology, Sydney, Australia

[ The player itself is named : RISCDEMO.ZIP. ]


===========================================================================
 XII | QUESTIONS
================

These are some questions, ideas or whatever problems, where still no
solutions is found or nobody knows an answer. Please contact me via e-mail
if YOU find a solution for:

1) Is there somebody out there (maybe from the MPEG-group), that could
   rewrite the first section of this FAQ, to bring it up to MPEG-II ?

2) Is Xing connected to the internet or compuserve or something ?
   Is there somebody responsible for there product ?

3) Are there multimedia-specialized mailboxes out there ? Please send
   a filelisting of your mpeg-archive, a description of how to obtain
   the files, costs, connection times, telefon-numbers etc.

4) Is there a COMPLETE mpeg-stream somewhere out there ? Meaning including
   a system-, video- and audio-stream ...

5) Who can post something about the MPEG-card from Optibase ?
   Test ? Compatibility ? Prices ?

Please mail to:

  [e-mail]

if you have for inormation, than I have.

===========================================================================

The end of ...

        THE MPEG-FAQ
        ======================================================
        PHADE SOFTWARE Leibnizstr. 30, 1000 Berlin 12, GERMANY
        Inh. Frank Gadegast            Fon/Fax: [phone]

        [e-mail]

===========================================================================