Video conferencing quality can make or break an implementation project. If users have negative first experiences...
with the new technology, adoption can be affected significantly, wasting your substantial video conferencing investment. The underlying network infrastructure is important to video conferencing quality; in general the higher the experience quality, the more network bandwidth you will need to support it. So if you jump to the conclusion that high-definition video conferencing is the only way to go, be prepared to justify some impressive network infrastructure costs, or to explain why the video conferencing experience doesn't live up to users' expectations.
Before you tackle the video conferencing quality issue and invest in equipment, services or any necessary network upgrades, make sure you take the time and trouble to understand your overall video conferencing requirements, choose the type of video conferencing that is most suitable for your company, do your homework to choose the right video signaling infrastructure, and locate your MCU where it will most cost-effectively scale. Chances are that fiscal constraints won't allow you to "have it all," so you must harmonize your priorities with your budget to zero in on the right technical solution for your situation.
Understanding video conferencing quality
When it comes to video conferencing quality, the first major decision you will face is to define the video quality your users need. Two key factors affect video image quality: resolution (number of bits in the image) and frame rate (number image refreshes per second). Boosting either resolution or frame rate increases the network traffic burden. Although video compression algorithms do a good job of reducing bandwidth consumption, as the table shows, they can only do so much.
For example, a High Definition (HD) 720p image that refreshes at 30 frames per second (fps) creates about 664 Mbps of data (1280 x 720 image size x 24 bits of color and intensity information x 30 fps). This 664 Mbps is typically compressed down to between 2 and 4 Mbps for transport over the network. The table shows typical bandwidth requirements for the resolutions and frame rates shown.
|Transport bandwidth required||Resolution||Frame rate|
|128 kbps||CIF||15 fps|
|384 kbps||CIF||30 fps|
|512 kbps||4CIF||15 fps+|
|768 kbps||4CIF||30 fps|
|1 Mbps||HD720||15 fps+|
|2 Mbps||HD720||30 fps|
|3 Mbps||HD720||60 fps|
|4-6 Mbps||HD1080||30 fps|
Resolution and bandwidth requirements
How much resolution do you need for adequate video conferencing quality? A good way to think about resolution is as "pixels per face." A primary benefit of video conferencing is that you can see other participants' body language -- particularly their facial expressions -- so you need sufficient resolution to see facial detail. The resolution required for a room-based environment with four to six (or more) people in one camera image is much higher than for a personal video conferencing environment with only one person in the camera image. Personal video conferencing often includes a head-and-shoulders image of one person, with the face consuming 20% or more of the image real estate. CIF (Common Intermediate Format) resolution (352 x 288) would deliver about 20,000 pixels per face in this case.
In contrast, a room-based system with four participants is likely to display a much broader view of a conference room, and an individual face may only cover two percent of the screen. In this situation, an HD 720p image will deliver about 18,000 pixels per face. Although pixels per face are roughly equivalent, the room-based system needs to deliver two to four Mbps of bandwidth, whereas the personal system only needs to deliver 512 kbps to achieve the same result.
Frame rate and bandwidth requirements
Frame rate determines how often the image will be updated (frames per second) and thus affects video conferencing quality in terms of how well the video system displays motion. Video systems can often be configured to use a higher pixel resolution by dropping the frame rate. Although this tradeoff can work well for static images, it degrades quality for dynamic images of people moving and talking. Slow frame rates deliver jerky video images that detract from the conference experience. So if bandwidth is limited, it often makes sense to ensure a full 30 fps, even at the expense of some resolution.
Video resolution settings should also take content sharing into account. On a PC desktop, most systems will share resolution equivalent to the screen resolution of the PC. But in a video conference room, resolution may need to display a range of detail. For example, less resolution is needed for PowerPoint presentations with big font sizes than for detailed architectural or engineering diagrams or spreadsheets.
Most of the above discussion assumes video is being used to connect participants in a conference. If video conferencing is being used for clear images of small details, resolution and motion support must be sufficient to make the application work well. It's a good idea to try the application in situ to ensure the resolution and motion handling will deliver the needed results before you buy.
About the authors:
John R. Bartlett is a principal consultant at NetForecast, where he focuses on network support for voice and video conferencing. NetForecast provides consulting to enterprises and networking equipment vendors on application performance issues and convergence of voice and video conferencing on the IP network. John has 32 years of experience in the semiconductor, computer and telecommunications, and has been consulting since 1996. John can be reached at firstname.lastname@example.org.
Rebecca Wetzel is a principal of NetForecast and a networking industry veteran with unparalleled inside knowledge of the network service and product markets. She works with network product vendors and service providers to develop and implement product strategies. She can be reached at email@example.com.