During the dark ages of the early
1990’s, people waited many minutes to download video clips from
the Internet. Finally, someone thought, "Why don’t we watch
one part of the video while the rest of it downloads?" Thus
streaming video, a system of special compression and buffering
techniques was born. It allows you to transmit real-time video
over the Internet to an exponentially expanding worldwide audience.
You may ask yourself "What
does this have to do with me? I’m not a computer person."
But if you’re a videographer, you might be interested in a way
to present your work to a potential audience of millions. After
reading this article, you’ll understand:
- what streaming video is,
- how to shoot video for streaming,
- how to digitize and compress the video onto your hard drive, and
- why you need a streaming server.
So keep reading and step into the
future of video.
To view a streamed video file from
the Internet, you need to install a plug-in (special helper software
that works in conjunction with your Web browser to handle the
downloading and decompression of the video). To perform the miracle
of transmitting and receiving video and sound through a phone
or high-speed data line, streamed video uses a nearly magical
formula of compression and buffering.
To compress video, a complex mathematical
formula breaks the video into individual frames. Each frame is
broken into moving and static components. Compression software
takes each moving object and guesses where it will be in the next
frame. By refreshing only the moving components of a frame, and
recycling the static, compression reduces the size and transmission
time of the video file. Similarly, in The Flintstones,
the animators saved countless hours by drawing only the moving
parts of each frame. Fred’s mouth was often times the only thing
moving in a scene, thus that was the only part actually redrawn
for a frame, the rest was recycled from the previous frame. There
is a downside to compression. If the camera is panning, zooming,
or moving in any way, the whole image is in motion, leaving nothing
to recycle. This leads to poor compression, and slower transmission.
Compared with the voodoo magic of
compression, buffering seems fairly bland. By giving the file
a few seconds to load before starting to play, a reserve of video
is available in the memory of the viewing computer in case the
transmission slows for a moment. When the transmission slows,
the viewing computer uses part of the reserve of video. If the
transmission becomes too slow, the buffer completely exhausts
itself, and the video plays at a lower frame rate and becomes
choppy. Better streaming packages will adapt the buffer size to
the speed of the connection, even as it fluctuates with the normal
ebb and flow of Internet traffic.
Shooting for Streaming
Streamed video is heavily compressed,
so your goal is to maximize the quality of the video, while minimizing
the amount of digital artifacts and choppiness that will inevitably
occur in a streamed video.
Remember that if the camera is moving,
everything in the shot is moving too. To keep your streamed video
from slowing down, shoot images with very little motion and use
a tripod. Avoid high-motion shooting techniques, such as panning
and zooming. This is not to say that you must eliminate all motion
in your shots, but try to make sure that the only thing moving
is the main subject.
The colors you choose can also affect
how well your video will compress. Bright solid colors are best.
Dark colors can confuse the compression software by blending together
with subtle shadows. The contrast of colors is crucial too. Choose
subjects that contrast distinctly from their backgrounds. Patterns,
either in the foreground or background, must be constantly refreshed
when there is the slightest movement.
Digitizing and Compressing
If you use an analog camcorder, you’ll
need a capture card to digitize the video. A video capture card
is an expansion card that works in conjunction with, or replaces,
the existing graphics adapter inside your computer. If you used
a DV (Digital Video) camcorder with a FireWire jack, you’ll be
able to transfer the video directly onto your hard drive with
a FireWire capture card.
With analog video, the first step
is compressing the video into an AVI, MOV, or MPEG format to store
on your hard drive. Compress the video at the lowest available
ratio that allows the clip to fit on your hard drive. Don’t worry
about the file size at this point, because you will have to compress
the video again into a streamable format.
Once your video is captured, edited
and rendered into a file onto your hard drive, you need to select
one of the streaming software packages to further compress the
file into a streamable format. The software you choose needs to
be supported by the streaming server you select. Find out which
software packages your prospective server supports before purchasing
software or hosting services (See sidebars of streaming software
and streaming servers).
When you compress the MPEG, MOV,
or AVI into the streamable format you have to choose exactly how
you want it compressed. The first decision to make is how much
of the video quality you want to sacrifice to make the video play
better over the Internet to phone-modem-connected users. The three
factors that make up the quality of a video are frame rate, color
depth and resolution.
Frame rate is the number of still
images that make up one second of a moving video image. At 30
frames per second (fps), images seem to move fluidly and naturally.
Video digitized at a frame rate of less than 15fps becomes noticeably
jumpy. Since current phone and modem technology limits the frame
rate to 10fps, you might as well set 10fps as the target in your
compression scheme. If a video clip contains a lot of action,
then limit it to 7fps.
An alternative to using 7fps streamed
video is the Web slideshow, which limits the frame rate to one
frame every five seconds. This allows for one higher resolution
image to be displayed with an audio track, instead of 35 low resolution
frames. (See the May 1998 issue for more about Web slideshows.)
The second quality variable is color
depth, the number of bits of data the computer assigns to each
pixel of the frame. When there are more bits of data assigned
to color each pixel, there are more colors that can be emulated
on the screen. Most video is either 8-bit 256 color, 16-bit 64,000
color, or 24-bit 16.8 million color. 256-color video is very grainy
and unsuitable for viewing. 24-bit color is optimal, but because
it greatly increases the size of the streaming file, you’ll want
to settle for 16-bit color.
The third factor in determining the
visual quality of the picture is resolution, measured in number
of pixels. The more pixels there are in your picture, the higher
the resolution of the video. For example, if your video is 640×480,
you have 640 pixels across each of the 480 vertical lines of pixels.
Streamed video ranges in resolution from postage-stamp size (49×49
pixels) to 160×120 pixels, the highest resolution you can stream
to users connected through a telephone line, all the way up to
640×480 and beyond, which is considered full-screen video.
Serving Up Your Video
To provide streamed video on your
Web site, you either have to serve it from your own Web server,
or have it served from a remote video streaming service provider.
Since few of us have own Web server we’ll avoid the techno mumbo-jumbo
and just cover what to look for in a streaming server.
The first consideration is how many
"streams" a remote streaming provider will allot to
you. A single stream, for example, would allow one viewer to watch
one video clip at a time. If you had 20 minutes of streamed video,
10 streams would allow up to 720 people to watch that video in
a 24-hour period. Some servers charge you by the number of streams
you want available to potential viewers, while others charge
you by the number of streams actually served to viewers.
Some servers do not sell their service
by the number of available streams. Instead, they charge by the
Megabyte for the total amount of your data that they have served,
regardless of the number of concurrent streams that your viewers
Another thing to keep in mind when
choosing a streaming server, is that not every company will serve
every format of streamed video. When you’ve chosen a streaming
software package that you want to use, make sure the serving company
you are considering can accommodate it (see sidebar).
Streaming video is just in its infant
stage, yet is has a potential limited only by the amount of data
that can be transmitted at one time, or bandwidth. However, bandwidth
is expanding exponentially. For the Web surfer, this could eventually
mean 200 million channels of TVesque video. For makers of video,
this could be an unprecedented broadcasting opportunity. For more
information about streaming video, visit Videomaker’s video streaming
FAQ at www.streamingvideos.com.
Larry Lemm is Videomaker‘s
[Sidebar: The Streaming Contenders]
Even though streaming video is fairly
new, there are already a few major players in the streaming software
RealVideo: RealVideo software
by RealNetworks (www.real.com) is the current leader of
the video streaming industry. Its browser plug-in and encoding
software are free, but you have to pay for the server software.
Because of its large installed base, RealVideo is a solid choice
for any potential video streamer.
Netshow: Netshow (www.microsoft.com/netshow/)
is an up and coming video streaming package from industry giant
Microsoft. The plug-in and encoder are free, while the serving
software is compatible with RealVideo, Vivo and VDO to allow a
video server to serve a variety of streaming formats.
VDONet: VDO by VDOCorp (www.vdo.net)
was one of the earliest streaming packages, but is now losing
ground in popularity to the other formats. The plug-in is free,
but you’ll have to pay for the encoding and serving software.
[Sidebar: Streaming Servers]
Here is a list of companies that will host your streamed video.
Video 2 Net
911 Media Arts Center