Creative Resources, Online Creative
Services, Design, Photography, Audio and Video Production
Using Online Audio and Video — some alternate
deployment models
by Rick Hendershot
Online Audio and Video Resources
(This article was
written in 2002. Not much has changed in the online audio and video
world since then. Although there has been relatively rapid deployment
of broadband connections across North America, commercial applications
of online media are still stalled, and apparently going nowhere. Why
this is the case will be the subject of a future article. One other
fairly significant change has been the widespread use of Flash, and
the development of viable
Flash video encoding. This should eventually have an impact on the
more widespread use of online video-on-demand. But don't hold your
breath. -- October 31, 2004)
For several years now, companies and organizations
around the world have been gradually turning to "streaming media" on the internet to
communicate their unique messages, and provide their clientele with precise
information about their products and services. In this presentation I want
to touch on some of the typical scenarios in which streaming media is
being used. These scenarios are what we might call "deployment
models". I will briefly discuss how these models work; and I will
touch on some of the advantages of each model, as well as its
shortcomings.
|
Recommended Creative Resources
Videoinabox - We don't
create online media (not much anyway). We just talk about it.
Creative
Colour Vinyl Banners and Large Format Posters - We produce
full colour vinyl banners from your digital files, and ship them
across North America. All vinyl banners are printed with outdoor inks.
Our prices are the lowest on the web. We also produce
popup displays, banner stands, retractable displays, and trade
show graphics.
Patriotic
Vinyl Banners - Put your patriotric graphics on a vinyl
banner. We have a library of free patriotic graphics. Vinyl banners
for events, rallies, trade shows. Full color graphics. Outdoor
printing. Lowest price on the web. Vinyl banners shipped across North
America. |
Model #1: The
Video-on-Demand Archive
On first impression this looks like the easiest model to implement from
the technical point of view, and the one that is most consistent with our
now familiar view of the internet as a source of information. Media files
are encoded in streamable formats and then they are put on a media server
as a resource that can be accessed via hyperlinks in web pages.
Consider, for instance, a
catalogue of movies with hyperlinks to the actual movie files. You click
on the hyperlink for "The Matrix", and the movie is streamed to your
computer where it plays in the appropriate player.
This model can be applied
to any "catalogue" situation. Imagine a university that has developed a
series of videotaped lectures, or a software training company that sells
instructional videos for specific software packages. It is relatively easy
to encode these lessons, place them on a server, and make them available
to students or subscribers on a "view-on-demand" basis.
This is also the typical
model used for product description videos. Say your company has five main
products — five different kinds of widgets. You make a video for each
widget, and anyone interested in a specific widget can just click on the
link and see the appropriate video.
The problem with this
model is that it only works fine as long as relatively few people are
accessing the media. Since every viewer requires a unique media stream,
the bandwidth used by a busy site can become excessive — and very
expensive. Once a particular service becomes widely viewed, it becomes
difficult, if not impossible, to recover the bandwidth costs involved in
servicing that audience.
I refer to this as the
BUR problem, where BUR stands for Bandwidth Usage Redundancy. Think of
what might happen when 50 or 500 or 5000 different viewers watch "The
Matrix" all at slightly different times — they are all, hypothetically at
least, viewing their own unique stream, and the bandwidth redundancy would
be tremendous. Consider how much bandwidth would be saved if they could
access the same stream at the same time — like watching a traditional TV
or Radio station.
Such considerations
inevitably drive the on-demand model in two different directions.
You can likely imagine what they are, and I discuss this at considerably
more length in my presentation entitled "Channels".
Model #2: The Live
Webcast
This is what many people first think of when they hear about streaming
media — the live webcast — because they have heard about live rock
concerts being broadcast on the internet, or phone in shows where some
celebrity or politician takes a few live calls while broadcasting an
interview over the web.
Webcasting has a certain
amount of appeal, because it seems to give almost anyone the ability to
broadcast their concert, annual meeting, conference key note speaker, or
golf tournament award ceremony to virtually the entire world.
From a technical point of
view setting up a webcast is more complicated and somewhat more risky than
creating a library of archived files. The first complication is that it is
a live event, and so involves all the usual risks associated with live
events. Things usually don't go quite as planned; equipment breaks down;
cables are forgotten back at the studio; lights burn out, people say
stupid things or start drooling, etc., etc., etc.
The second complication
has more to do with the specific technology used in webcasting. Since you
have to "encode" a media stream before sending it out over the net, the
encoder has to work more or less in real time. The cameras and audio
equipment feed the signal into the encoding device. Then the encoding
device has to compress and format the data so it can be sent along to the
media server — presumably at a different location — where it is then
streamed at the correct bitrate and ultimately played by the correct sort
of media players at the other end.
But of course there is no
"correct bit rate". Some people connect with 100k connections, and others
with 300k connections. And most streamed signals are made available to
people with different media players — Windows Media, Real, or Quicktime.
Doing this with a webcast is not practical. So the webcaster just has to
pick what he or she considers the format most likely to be used by the
largest number of viewers, and go with it.
The third complication
has to do with getting the signal from the remote site back to the media
server site. How do you do this? If the remote site has a reliable high
speed connection that can be more or less monopolized for the live event
then perhaps it is feasible. But if that isn't available, the webcaster
will have to join the big leagues and go for some dedicated link purchased
from one of the usual data haulers: a dedicated phone or cable link, or
maybe even a satellite hookup.
Is this likely to be
expensive? Yes. Does that substantially reduce the number and kind of
events that can be webcast. The answer is obvious isn't it? So is event
webcasting really in inexpensive way for your average garage band to make
their mark on the world? Not likely. As with most sophisticated, high tech
communication options, the only users will be those who can afford the
technology. And generally they'll only be willing to afford it if there is
a satisfactory return on their investment. So webcasting ultimately just
becomes another technology that only large corporations can use. And they
will use it to push the same old messages we've been hearing on all the
other mainstream media.
Perhaps there are a few
exceptions to this. A few months ago I heard of a company setting up
cameras in daycare centres and webcasting video of the children so their
parents at work could watch their children on their computers.
This idea may or may not
ultimately fly; but it suggests a range of possibilities from video
surveillance to in-house newscasting, to narrowly focussed corporate event
coverage, where we are not talking about one-off events like concerts, but
rather regular coverage of recurring events in a controlled setting — like
in a studio or newsroom, or briefing room. In such a situation the
broadcast source would be controlled and predictable, the transmission
system properly constructed and tested, and the viewing audience motivated
to watch and equipped with the appropriate tools to do so.
For other articles in this series, go to
Online Audio and
Video Presentation Models.
| 