| ★ wanayoo — archive 1999 http://www.samizdat.com/video.html | Nouvelle recherche | Portail wanayoo |
Permission is granted to make and distribute verbatim electronic copies of this article for non-commercial purposes provided this permission notice is preserved on all copies. All other rights reserved. To correspond with the author, send email to: seltzer@samizdat.com
I am not a great fan of multimedia on the Internet. I don't go for flash for the sake of flash. I keep looking for the applications that justify the expense. And, let's face it, up until now the audio and video on the Internet has been a sideshow -- a curiosity, something you demo to newcomers, but that you don't really use.
That seems to be changing. RealAudio (http://www.realaudio.com/) is easy to use, and the sites being created by their customers becoming increasingly tempting. And now VDOLive looks promising on the video side (http://www.vdolive.com/) . (See comments above).
VDOLive seems to have the same sound quality as RealAudio, only you see pictures too (cf. MTV).
RealAudio seems designed for pre-packaged content (with the neat feature that you can embed URLs in the audio stream and hence create slide shows or something analogous to the film strips that were used in schools back in the 1950s, with the pictures in synch with the sound).
VDOLive seems like it can be used for live broadcasts as well as later playbacks.
With RealAudio the emphasis is on the "real" as in "real-time" -- the fact that you start to hear the sound shortly after you click (not having to wait for the file to download). VDOLive is also "real-time," but the emphasis seems to be on "live." It looks like you have here the opportunity for live broadcast interviews, in addition to prepackaged content. With poor video resolution, this medium feels better suited for programming that looks live and spontaneous rather than edited and polished.
On the server side, RealAudio lets you set up the equivalent of a radio station, without a government license. Now, appearing on the market just a short while later, we see VDOLive providing the rudiments of what could let you set up your own basement television station. This would be tempting stuff to play with on the server side, if you could afford expensive toys.
But the "live" aspect also points toward important, business-related applications.
Surprisingly, two-way audio/video videoconferencing software isn't market-ready yet. CU-SeeMe from Cornell (ftp://gated.cornell.edu/pub/CU-SeeMe) is available with two-way video and audio for the Macintosh, and just video (no sound) for the PC. But it looks about as user-friendly as a Heath kit for building a bomber.
The promising features of VDOLive seem to indicate that clever compression techniques, line speed -- even at just 14.4 -- might not be the major barrier it looked like a year ago . In other words, we don't have to wait until nearly everyone has through ISDN lines or TV cable lines in order for useful videoconferencing to become widespread. And I'd be very surprised if we don't start seeing products of that kind from professional, commercial sources, rather than university labs, very soon.
When video and audio are live and two-way, they can serve as a means to extend the ways thatpeople relate to one another, and they also can become very efficient means for communication -- conveying all the non-verbal information that is essential in negotiating and building relationships.
In Neal Stephenson's cult-classic novel Snow Crash, it is the ability to recreate subtleties of facial expression in "avatars" (on-line alter egos) that makes virtual/alternate reality take off.
I suspect that there will be three stages in the development of video for the Internet -- tied todevelopments in visual quality (resolution) and speed (frames per second). What we have now is little better than mailing somebody a still photo, or having a still photo on the screen while someone talks (for instance, with Internet Phone). Yes, there's a recognizable face, but there's no connection between the expression on that face and what the person is saying.
The first stage will probably be the achievement of natural-looking movement (through even greater compression or just widespread availability of high bandwidth at low cost), so gestures seem natural, and the person at the other end seems more like a human being than a poorly programmed robot.
Next will come the very high levels of resolution which are necessary to capture all the subtleties of human expression. (If we were going to be limited to today's speeds, I suspect that detailed analysis a myriad of facial expressions could lead to low-bandwidth approximations -- focusing on just those aspects of the face that seem to convey the most nformation about a person. But we won't be limited, and hence that kind of research isn't necessary.)
But the real winner -- the Holy Grail of personal videoconferencing -- will be eye contact. How do you (given that you have solved the above two issues) mount and synchronize the cameras to create the compelling illusion that you are looking someone in the eye and that person is staring right back at you? Then you will be able to communicate a full range of emotion, with all the associated messages of sincerity and credibility; and video over the Internet becomes a major, indispensable, business tool.