The explosion of orders for videoconferencing systems is a clear indication that remote meetings are becoming a “mission critical” tool in our post-9/11 world. But whether the videoconferencing star remains bright will depend in large part on organizers who can produce an effective videoconferencing experience. Users need to take the time to understand how videoconferencing works in order to understand its limitations and its potential. A good place to start is with a piece of equipment called a codec.

Compare and Contrast

The job of the codec (which stands for encoder/decoder) is to compress the constant stream of signals from the video cameras and microphones. Once compressed, the signals can be sent over ISDN circuits to the far site without using an unruly amount of bandwidth.

As soon as the connection is made, the codec encodes the image sent to it by the video camera, writes it into memory, and sends it off to the far-site codec. This first image is called the “key frame.” The far-site codec receives the key frame, writes it into its memory, and displays it on the far site monitor. Then, the sending codec compares subsequent images (30 times per second) to the original and notes which parts have changed. The sending codec only transmits the parts of the image that are new, and the receiving codec pastes those changes into the image in its memory and displays the composite image. The bottom line: Only motion is transmitted. With this clever technology, bandwidth requirements are minimized because the static parts of the image are not resent.

This is a key concept to understand for a successful videoconference, because the less things move in the image, the smoother the motion looks to the viewer. When the camera sends a lot of new information to the codec, the codec sends out a new key frame. When this happens, the connection circuits are jammed with data, and the codec bogs down, sometimes transmitting as slowly as one frame every two seconds as it struggles to keep up with the flood of changing information. Of course, this dramatically reduces the effectiveness of the conference and can be quite distracting to the viewer.

Don't Move a Muscle

What looks like motion to a video camera? Well, obvious motion such as participants rocking or swiveling in their chairs. Solution: chairs that don't move around as much. Another factor that can be controlled is the panning, tilting, zooming, and focusing of a remote-control camera. This can be addressed by programming the system and educating the operators. What if the camera can “see” an open door or hallway where people walk past? That's motion too, and the solution is obvious: Give the camera as much of a fixed, unchanging image as possible.

Let's consider less obvious motion. What if the camera is vibrating a little bit? Yep, that's motion. Some office buildings have “post-tension” floors that vibrate and shake when people walk across them. If a stable room or floor cannot be found, you'll need to reduce traffic near the meeting room. I'm unaware of gyroscopically stabilized VTC cameras or lenses that could help solve the problem.

Are there windows in the room? Even clouds passing in front of the sun and changing the light levels and colors in the room appear as motion to the camera. Try blackout curtains on all windows, or use an interior room.

Is the camera's auto-iris turned on? Each time the camera adjusts to changing light levels from papers moving around, shadows, and so on, the codec has to build new key frames. Solution? Get good lighting (ideally 80 percent reflected), and turn off the auto-iris.

By understanding how the codec reacts, we can make its job easier, and the quality of the videoconference will be far less distracting. We'll address audio issues in the next issue. In the meantime, e-mail me your questions and ideas for future articles.




Jeff Loether is founder of Electro-Media Design Ltd., a Rockville, Md. — based technology consulting and design services firm. Contact him at (301) 309-0110, jloether@electro-media.com, or visit www.electro-media.com.

Videoconferencing Systems: Three Tiers

  • Desktop systems use a computer and a small monitor-top video camera, are typically for internal desk-to-desk communications, and are “IP” based, meaning they use the enterprise LAN and can communicate with other IP-based systems outside the enterprise via the Internet.

  • Portable (or roll-about) systems have one or two video monitors and, commonly, one camera on top of the monitors. They may have other accessories such as a document camera, electronic whiteboard, scanner, printer, or second room camera. These elements are usually assembled on a cart to be rolled to various rooms as needed. And typically these system connect to digital ISDN telephone lines; three ISDN lines is considered the “business standard” connection rate.

  • Built-in systems are ordinarily found in board rooms, distance-learning facilities, and other dedicated videoconferencing rooms.