“The videoconferencing use model is going through a tremendous transformation. The decades-old conference room model is quickly giving way to more effective solutions. The telepresence room has demonstrated what is possible—high quality, low latency video that creates a natural, engaging interaction. Telepresence innovations from the old conference room solutions include:
Eye contact—Reorient people to face the screen and camera, making individuals the focus of each camera, not the conference table
Standard definition per face—Deliver high resolution on each person’s face so that even subtle reactions are understood
Consistent quality—The picture should never break
Not more than 250 ms of latency—Minimise latency to maximise real-time collaborative experiences, just like a standard audio conversation.
Telepresence rooms demonstrate one of the key principles of quality interaction—everyone is equal. The polar opposite of interaction comes from the audio conferencing experience where some people gather in a room and one or more call into the conference room audio device on the table. The interaction among those in the room is far superior compared to the frustrating experience of those who could not be in the room. The inability to see and be seen creates an unequal and unfulfilling interaction experience.
There are many factors driving people to use videoconferencing. Telepresence has shown that it can work effectively. Economic pressure to lower costs and improve productivity in the context of dispersed and flexible team configurations are driving the move to improve the quality of communication. Videoconferencing done right delivers on all of these motives.
Moving Telepresence to the Desktop
For widespread use, everyone who wants to use videoconferencing needs to be able to get access whenever they need it, wherever they are, just as they do today with personal communication devices for voice. And they need videoconferencing to meet the high quality threshold set by telepresence. The challenge is that telepresence rooms are very expensive to acquire and operate, so access is limited if available at all.
Gartner predicts tremendous growth in videoconferencing. Existing SD quality room system deployments will be replaced by telepresence and HD quality rooms, but overall the room system number will remain relatively flat. Gartner further predicts the explosive growth will be in desktop endpoints. For this adoption to take place, the desktop deployment must also deliver the high quality experience expected by users.
Gartner uses the example of a very large organisation videoconferencing deployment with 200 room systems and 50 desktops today. Within two years, this distribution remains at 200 room systems and grows to 1000 desktops, and continues to accelerate on the desktop deployment. This massive endpoint growth will require changes in the way videoconferencing utilises the IT infrastructure.
A New Architecture is Required
Most of the SD room systems are obsolete today and don’t meet the usability quality threshold, so they will be rapidly replaced. In fact, many laptops today can deliver videoconferencing experiences that exceed those of the obsolete SD room systems. But moving to HD quality with the same solution providers only puts more pressure on the network to provide more bandwidth at high service levels.
The opportunity for IT is to adopt an architecture that removes the requirement of expensive high QoS networks for video, while also enabling the ten-fold growth in endpoints. This will deliver access to videoconferencing for teleworkers and road warriors, as well as those who work in the office. The new architecture should require only a best effort, low cost network to make this option affordable in the face of much greater demand for bandwidth.
It is not economically viable to support the massive increase in multipoint videoconferences via the Multipoint Conferencing Unit (MCU) architecture. Legacy videoconferencing solution providers use costly hardware to transcode video from different sources into a form that can be used by each endpoint. Rather than benefiting from Moore’s Law the way the PC industry does, the legacy videoconferencing solution providers require increasing costs for increasing quality without the benefit of economies of scale. Beyond the sheer expense, the fundamental flaw of the decades-old architecture of a multipoint room-based deployment architecture from legacy videoconferencing solution providers is that it fails to deliver the telepresence experience required for large scale adoption.
So if the architecture of the past will not support the deployment model of the future, what can you do to prepare your network for the growing demand for desktop videoconferencing? The answer is in a new videoconferencing architecture built upon the newest video compression standard—H.264 SVC—enabling the removal of the high QoS network requirement while maintaining the low latency, high resolution per user face quality threshold required for mass adoption.
The VidyoRouter™ architecture is designed to connect large numbers of videoconferencing endpoints over best efforts networks—including the Internet. The VidyoRouter architecture eliminates transcoding and routes packets individually to each endpoint as required, providing the best experience for every user. The result is an IT infrastructure that optimises bandwidth use, minimises capital investment and lowers operating costs.
In addition to being able to use low cost bandwidth, VidyoConferencing™ further lowers costs by efficiently using HD conferencing ports and minimising bandwidth consumption with the following architectural and licensing advantages:
Floating port licenses enable reuse of HD conferencing ports wherever they are required. Ports can move to whichever VidyoRouter requires them, enabling a global company to have VidyoRouters located in each region. The maximum number of ports needed is equal to the maximum number used at any time. This removes the duplication of hardware ports that occurs with an MCU-based architecture that becomes very expensive as the number of endpoints rises. And since the ports are not tied to specific appliances, redundancy can be achieved easily and inexpensively by adding low-cost VidyoRouters. With VidyoConferencing there are no wasted ports in any location, and the total number required is dramatically lower than hardware based solutions.
Since VidyoRouters can be deployed to each continent to provide localised network access, endpoints can utilise the VidyoRouter closest to them so latency is further reduced and bandwidth is conserved.
In multiparty meetings that span multiple VidyoRouters, eventually inter-VidyoRouter traffic will be consolidated to only a single stream between VidyoRouters—with local distribution to each endpoint—so that long-haul traffic will be minimised.
Since the VidyoRouter has software-based ports running on Intel x86 hardware platforms, it enjoys the same performance growth curve as the PC industry and currently supports more than 100 concurrent HD multipoint connections on a single 1U appliance (about 10X density advantage over traditional MCU while providing rate matching and continuous presence).
The overall benefit is that as the number of endpoints expands, both infrastructure and bandwidth savings grow over the traditional MCU architecture.
VidyoConferencing Delivers Telepresence Quality Across the Entire Product Range
Comparing usage patterns of Telepresence and legacy videoconferencing shows that a minimum performance threshold is required for user adoption. The natural interaction of Telepresence drives high usage rates, but the costs and limited access prevent widespread deployment. VidyoConferencing delivers at least SD quality per face with the required natural interaction, across the full range of endpoints. And with today’s PCs and broadband networks it is possible to deliver Telepresence quality to the desktop—Personal Telepresence.
Organisations will be driven to provide videoconferencing on laptop computers, and this will result in an exponential growth in video enabled endpoints. IT will find that this requires a new architecture that is capable of consistently delivering Personal Telepresence to the desktop and addresses the issues of quality, affordability, accessibility, and scalability. As discussed above, traditional MCU-based architecture can’t deliver the quality users demand for adoption and doesn’t scale easily or cost effectively, as MCU cascading and QoS enhanced networks to many remote endpoints is both limiting and very expensive. Moving now to a new architecture that consistently delivers the telepresence experience over general purpose IP networks and scales as naturally and cost-effectively as the Internet itself, will position your organisation to reap the productivity and cost saving benefits of ubiquitous video enhanced communication. Vidyo’s architecture can be ten times cheaper to deploy and vastly more scalable while providing better user experience."