TelePresence Audio and Video Technologies

Codec Design Requirements

When Cisco began its quest to design and build a true-to-life telepresence experience, it had a rather large obstacle to overcome. No other vendor at that time had anything that could deliver the level of video and audio quality and integrated systems approach that Cisco wanted to create. So Cisco decided to build the codec from the ground up, purpose-built to meet the following requirements:

• First, it had to deliver multiple channels of 1080p resolution video, at a consistent 30 frames per second, at extremely low encoding and decoding times and at a bandwidth utilization rate the average large enterprise customer could afford to deploy.
• Second, it had to provide multiple, full-duplex channels of wideband audio. Furthermore, those channels could not be mixed together, nor stereo (left/right), but had to be discrete, independent channels to preserve the spatiality and directionality of the audio.
• Third, to meet its ease-of-use and reliability ideals, it had to provide a fully integrated system that could be completely managed by a single interface. Therefore, it could not utilize off-the-shelf components, such as cameras and displays, but needed integrated components that were managed and controlled by the system.
• Fourth, it had to provide integrated, easy-to-use data collaboration and audio conferencing so that participants could easily share documents and presentations with each other and allow audio-only participants to join the meeting.
• Finally, it had to do all these things in a completely standards-based way, leveraging the existing converged IP network and Unified Communications platforms Cisco was famous for.

As luck would have it, the time was right for the development of such a product:

• Digital Signal Processor (DSP) technology was just beginning to be capable of 1080p resolution video at the latency targets required.
• Camera lens and sensor technology was getting small enough to deliver a 1080p camera that was small enough to be discretely mounted over the bezel of the display.
• Display technology in the size (65-inch) and resolution (1080p) required was becoming affordable.
• The networking technologies required to make a system like this manageable and deployable on a converged IP network were finally mature—IP Telephony being one of the driving forces in the decade prior to Cisco TelePresence to allow technologies such as quality of service (QoS), Power over Ethernet (PoE), high availability, and Session Initiation Protocol (SIP) to become mature, allowing the same techniques to be leveraged by Cisco TelePresence.

The word codec is used in two different ways:

• It is used as the name of the physical device containing a CPU, memory, an operating system, Digital Signal Processors (DSP), and audio, video, and network interfaces: the Cisco TelePresence codec.
• It is widely used throughout the industry to refer to various audio and video encoding and decoding algorithms, such as the H.264 and AAC-LD codecs.