---

6 months ago I wrote an “Ask the EXPERT” post on Scalable Video Coding (SVC) and the future of video conferencing. This was following a good comment on another post, discussing the ever-growing bandwidth requirements of video applications. As our industry is in a constant chase after performance, bandwidth, picture size, you name it, scalability becomes a very sought-out feature.

As I’ve written in details in the original post, the scalability features of SVC, specifically of H.264/SVC, the recent amendment to the popular H.264/AVC standard, offers great possibilities. The idea of allowing the adaptation of a single video stream to different endpoint capabilities – performed by the endpoint or any network element on the way – sounds too good to pass.

But here’s the catch: if all this is too good to be true, why hasn’t scalable video coding caught on before? After all, it is not a new concept. Well, in that post I noted a few problems that prevented SVC from gaining commercial success:

• Previous SVC coding standards were not efficient enough to be productized
• No compatibility with legacy infrastructure (backward compatibility) and non-SVC equipment
• No standardization of SVC in application areas, which means no interoperability between the proprietary SVC systems.

The H.264/SVC standard, which gained a lot of market hype ever since it was ratified in November 2007, solves the first problem. However, even this new standard fails to solve the other problems, and to present a complete solution for the visual communications demands of the 21^st century user.

The Gateway Approach

SVC technology offers great benefits, including great error resiliency and the ability to serve different clients with the same video stream. However, legacy and non-SVC infrastructure can’t use it. Other SVC systems can’t use it too. How do you solve these issues?

A natural solution is the “Gateway” approach – add a gateway between the SVC silo and the rest of the world, this gateway will transcode SVC to non-SVC (AVC) and vice-versa, and you have a valid solution.

The SVC “Gateway” Approach: Transcoding

Drawbacks? Well, while the SVC silo enjoys the benefits of SVC, existing systems and non-SVC equipment will now get a degraded experience, as they have to go through the “Gateway” for every call, even if both parties who are not SVC-based. Such a call will be transcoded once, from AVC to SVC, and then transcoded again, from SVC back to AVC.

The result will be additional delay and degraded quality. So while you are trying to shoot for a scalable video solution, in reality this solution is not scalable at all, in terms of the various locations, systems and equipment a user can choose from.

The RADVISION Solution

In RADVISION we were well aware of the need for a truly scalable solution. But as a leading force in visual communications standardization, our main concern was to provide a solution that can truly interoperate with all equipment and infrastructure, a solution that is market based and not technology based.

True to our own motto of “don’t follow, innovate“, we decided not to embrace the “gateway” trend, but to adopt a an innovative approach, the one we announced officially today in the WR Collaboration Futures Summit in Berlin.

The visual communications market has evolved, in the past few years, from the meeting rooms to the desktops, the home offices and the high-end Telepresence suits. This versatile world indeed requires a scalable solution, but one that is truly scalable: one that not only can guarantee a high level of experience despite network problems, that are known to degrade quality; but also one that can connect any endpoint, be it legacy equipment, non-SVC equipment or new SVC-based clients.

RADVISION’s SVC-based Solution

RADVISION’s solution works with any device, without the need for additional components or transcoding, and without any increase in delay or decrease in quality. Through its new SVC-based technology RADVISION can offer full video error resiliency over networks with up to 40% packet loss, while maintaining high quality and low delay. Yes, no typo in the percentage of packet loss. It’s 40%.

RADVISION’s technology uses Unequal Error Protection (UEP) methods, designed for data protection over erroneous networks, to provide full protection through the SVC layered structure. With UEP important layers are fully protected, while other layers are less protected (or not protected at all). The UEP methods, including Forward Error Correction (FEC) and retransmission techniques, are either standard-based or are currently being standardized by RADVISION.

Quality levels in different packet loss rate scenarios:
Traditional H.264 systems vs. RADVISION’s SVC-based technology

The result is very impressive, as you can see in this brief demonstration :

More information on the technology itself can be found in the dedicated section on our website.

The great thing about this, other than the impressive results, is that Non-SVC based systems can enjoy the same level of error resiliency just as well. The RADVISION solution brings the benefits of SVC to everyone, and thus is truly scalable:

• If you’re using point-to-point calling, for instance from one SCOPIA Desktop client using this technology to another, you can enjoy a high quality of experience, even if network conditions or client capabilities are limited.
• If you’re connecting from outside of the corporate network, using the public Internet or dedicated lines, from the home office or any other location, using RADVISION’s technology will ensure a high quality connection, even when remote connection is problematic (packet loss, limited bandwidth, etc.).
• When deploying a distributed MCU architecture, where MCUs are deployed in different geographical locations (branches), this technology will enable a perfect connection between network elements, even if some of the paths are error prone.
• And, last but not least, when connecting a non-SVC based endpoint to a conference, perfect interoperability will be maintained, so that there is no need for any additional investment, in order to enjoy this scalable solution.

And there you have it: through innovation, through thinking “outside of the box”, through catering to market needs, through evolution rather than revolution, RADVISION is changing the way visual communication looks and feels. Ira Weinstein, Senior Analyst for Wainhouse Research, was commenting on this:

“SVC will significantly improve the experience of video collaboration especially over unmanaged networks like the public Internet. And because RADVISION provides technology to so many of the industry’s vendors, it will help accelerate the adoption and deployment of this important technology.”

In the upcoming months RADVISION will release the first line of its video conferencing infrastructure, SCOPIA, supporting this ground-breaking technology. I can assure you there is a whole lot to wait for.