Cell Relay Archive

Cell Relay Retreat>List Archive>month:1997-Nov> msg00125



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

Re: MPEG-2 VBR frame size traces

  • From: bannai@best.com (Vinay Bannai)
  • Date: 18 Nov 1997 11:34:29 -0800

In article <347119BE.F3F9E295@cps.msu.edu>,
Praveen  <amancher@cps.msu.edu> wrote:
>Hi All,
>
>I am in desperate need of some MPEG-2 VBR traces, in the form of frame
>size sequence(kb/frame) (not the binary MPEG-2 streams). This is for my
>grad
>thesis involving the characterization of MPEG-2 traffic using frame
>sizes at
>intermediate ATM nodes. If you are in the possesion of such a trace or
>have
>the knowledge of any public ftp sites or archives maintaining MPEG-2
>traces,
>please inform me. Thank you very much.
>
>Thanks in advance,
>Praveen.
>amancher@cps.msu.edu
>

Praveen,

We had worked on something very similar a year back. It involved traffic
characterization of the MPEG/MPEG2 frames over ATM network. I don't know
what you mean by VBR traces, but I am assuming you want the trace file. We
had worked on something very similar in which the trace file had the
following information for the video frames.

<frame sequence no> <size of the frame> <type of the frame -I,B,P>

We did not have the time stamp, because the video frames are generated at
regular intervals (around 1/30 of a second). I quit working for the
company several months back and I don't have access to the data anymore on
a computer. I have it on a tape. It should have traces for several
different movies (Dick Tracy, The City of Joy, Utopia, Bladerunner,
Alien). We chose these movies for the different characteristics they
exhibited. You might also want to try the following e-mail address and the
home page (the principal author on this paper).

http://www-leland.stanford.edu/~agrawals
agrawals@cher.stanford.edu

Hope this helps..
Vinay



-- 
Vinay Bannai                     E-mail: bannai@best.com