[AG-TECH] Fwd: A Lip Synch Problem Solution

[Markus Buchhorn]

For info. Interesting development out of OSU...

>Date: Wed, 31 Mar 2004 13:19:00 -0500
>From: Bob Dixon <dixon.8 at osu.edu>
>To: Megaconference: ;
>Subject: A Lip Synch Problem Solution
>
>There have been a number of discussions about the occasional "lip synch" problem in video conferencing.  The problem is that in some cases the video is delayed longer than the audio, so voices are heard before the speaker's lips are seen to move. Crudely speaking, this is caused by the fact that the video information is more voluminous than the audio information, and hence is delayed by overloaded components along the way, such as network devices and MCUs.  There is no inherent synchronization built in, so in a non-ideal world, things tend to become unsynchronized.
>
>There is usually no "knob" or setting one can manually adjust to compensate for this. So this leads to either just accepting it, or looking for more unusual solutions.
>At the recent ViDe conference in Indianapolis, this topic was discussed, and a new approach
>was suggested. I have investigated this new approach further and found that it may be a solution.
>
>The basic idea is that the audio needs to be delayed with respect to the video, so as to bring them back into synchronization.  It turns out that people can learn to do this with a little effort.
>All you have to do is move your lips before you actually speak.  This is a form of ventriloquism, and it has been taught for some time at the Edger Bergen Institute for Advanced Speech.
>
>In traditional ventriloquism, you shift your speech to another location. But in this case, you shift it to another time.  In mathematical terms, the first case deals with the spatial domain (often involving the spatial Fourier Transform), and the second one with the time domain (where it is described by the McCarthy Transform).
>
>I have been studying this new technique via a correspondence course from the EBIAS, and have achieved a modest degree of skill, as of April 1.  I can demonstrate it for you some time in a video conference, or maybe in next year's Megaconference.
>
>
>                                                           Bob
>
>
>-- 
>Robert S. Dixon, Ph D, PE
>
>Chief Research Engineer
>
>Ohio Academic Resources Network (OARNet)
>and
>Ohio State University, Office of the CIO
>
>Postal address:                  Office Telephone: 614-292-1638
>Office of the CIO
>Room 451 Baker Systems Engr
>Ohio State University            Fax:              614-292-7081
>1971 Neil Ave                    Email:        Bob_Dixon at osu.edu
>Columbus, OH 43210
>


Markus Buchhorn, ANU Internet Futures Group,       |Ph: +61 2 61258810
Markus.Buchhorn at anu.edu.au, mail: Bldg #108 CS&IT  |Fx: +61 2 61259805
Australian National University, Canberra 0200 Aust.|Mob: 0417 281429