Video switching has encountered two technology trends that together are changing tasks in the professional studio radically. Native signals used in video editing evolved from analog to SD (standard definition) digital, then quickly to HD (high definition) and several higher-bandwidth standards.
At the same time, video aggregation that once was characterized by point-to-point dedicated digital streams is poised to take advantage of packet-switching technologies, with Ethernet showing particular promise as a Layer 2 delivery mechanism.
Of course, video switching, as defined by the professional studio, is quite different from the packet switching technology used in the enterprise LAN, or even the WAN backbone. At the same time, modifications and extensions to core technology have resulted in carrier Ethernet technologies that are enabling a migration of Ethernet from the enterprise to carrier networks.
The advantage in Ethernet adoption has been a uniform standard resulting in higher volumes, thereby driving down deployment and operating costs. In addition, the emergence of 10Gbps Ethernet bandwidth per link and audio/video specific mechanisms, such as audio/video bridging (AVB), lay the groundwork for delivery, transport and switching of uncompressed HD content in studio environments.
Before HD signals became prevalent, professional studio mixers and switchers manipulated SD and composite video signals, hardware needs were significant but did not always define the state-of-the-art highest performance driven by routers in communications markets.
The slow but steady arrival of 3G HDTV, 3D HDTV, 4K2K, and Super Hi-Vision (Ultra HD) brings bandwidth requirement speeds to the tens of gigabits per second. This speed demand will be accelerated further by expansion of local, national and international HD productions and the anticipated move to 1080p and 3D broadcasts which will drive up storage and intra-box communications links in a studio environment.
Meanwhile, the effort to move the first stage of editing and mixing video content from centralized studios to remote digital TV vans has placed new demands on hardware to reach new milestones in low power and high integration. These two goals now exceed the demands for signal quality that will maintain uninterrupted broadcasts to the viewing audience(Figure 1 below).
Changes in site-based video switching have followed two related trends favoring denser switches. The arrival of IP gateways in video networks has made it possible in some network topologies for broadcasters to capture and transmit uncompressed HD video from remote sites.
Also, broadcasters seek to simplify overall transmission-system architectures through the use of denser, higher-performance switches that carry the potential of replacing bulky remote studio trailers that are often the size of semi trucks. Miniaturized video switching platforms could lead to reduced sizes, lower power and simplified architectures in remote transmission.
Figure 1: New technologies are enabling new architectural alternatives for the traditional video router function.
An emerging trend is the packetization of high-bandwidth traffic, a familiar trend that has played out in telecommunications and carrier environments. Digital video editing highlights some of the architectural demands of new video switching architectures. The video industry has evolved from analog input and composite traffic over RF cable, to emerging architectures featuring sophisticated control, scalability and embedded audio capabilities.
This is similar to some of the operations, management and performance monitoring functions enabled in carrier Ethernet technologies. While video switches still push performance and scalability on a larger scale than multiport Gigabit Ethernet switches, the chip-level requirements are sharing more in common with packet switches, particularly in integrated low latency routing and Quality of Service packet prioritization methods. In addition, precision timing protocols such the IEEE 1588 protocol allows packet synchronization and precision timing in the sub-microsecond range.
The need in shifting to an all-packet infrastructure for HD video is taking on greater significance given the migration to multi-gigabit and 10 Gigabit protocols. As in the case of telecommunication traffic, the shift to packet video will offer revolutionary consequences, something that has been recognized in recent years by specialty hardware vendors and the Society of Motion Picture and Television Engineers.
Products have surfaced offering interoperability with telecom and CATV packet traffic. The net benefit is simplification of carrier network architectures and reduced complexity and operational costs. Initial development work is being carried out on 10Gbps video protocols with a deliberate intent of mapping the higher-end standards into 10G, 40G, and 100G Ethernet.
In parallel with the SDI efforts, SMPTE also worked on a standard for Video Over IP, which has been realized in the new standards such as the SMPTE 2022 standard for video transport and Forward Error Correction. When a professional video studio elects to move to packet transport, it must either insure minimal packet loss by adopting effective end-to-end QoS strategies.
Alternatively, it must retain the ability to reconstruct dropped packets at the receiving end, through the use of FEC techniques such as those specified in SMPTE 2022. In any event, these packet trends favor the use of semiconductor switches and physical-layer devices from vendors familiar with QoS, network timing and FEC techniques, as well as traditional SMPTE SDI implementations.
As video architectures evolve, existing dedicated point-to-point architectures are likely to give way to packetized router technologies that resemble standardized, high-bandwidth Ethernet switches used in carrier and telecommunications applications. The push towards greater scalability and higher per/port speeds in the 3-10 Gbs range align with the technologies emerging in carrier Ethernet such as quality of service, timing, optical transport and FEC.
The new video switch requirements break from past approaches in the number of ports required, need to control jitter and return loss complexities at multi-gigabit speeds over greater distances. The latter concerns are driving chip-level architectural decisions such as standardized input signal equalization and on-chip diagnostics.
Some aggregation tasks for the modern video control engineer are simpler than they used to be, because real-time mixing and aggregation is more automated, thanks to the interplay of software and complex VLSI devices.
At the same time, however, the increased number of channels and the different types of video signals place demands on continuously monitoring the console for degradation or loss of signal, as well as more significant network failures. In effect, the console engineer in a video studio often must monitor signal quality regimes as multifaceted and complex as the network managers in telecom networks responsible for overseeing Service Level Agreements.
The new proliferation of SD, HD, and Quad HD represents the best and worst of times in terms of viewer options and network complexities. At the same time, similar network evolution and advances have already taken shape in carrier and enterprise environments. The result has been a new level of user services and value driven by semiconductor and networking building blocks.
Technology evolution and product introductions point toward packet transport migration utilizing well-defined SDI interfaces. The need to monitor and control the aggregation of this type of video traffic means that all video switching and transport topologies must become QoS and implement timing and streaming reservation technologies such as 1588 and AVB. Another enabler in the video realm will be tools to monitor and control real time issues.
Semiconductor companies can aid this transition by offering integrated support for device diagnostics, system diagnostics, network timing, and QoS prioritization. Choosing intelligent components in a video switch architecture can simplify the task of moving to packet transport.
But system-level design engineers at video OEMs, as well as mixing editors in the studios themselves, must share a responsibility of becoming more network-aware, both in terms of understanding how the multi-gigabit HD traffic of the future is moving to packet transport as well as how this packet traffic is monitored and controlled to insure lossless delivery of the emerging HD traffic types.
Juan Garza, video product marketing manager, for Vitesse has more than 17 years of experience in the communications and semiconductor industries. Mr. Garza has held various business and marketing positions at Texas Instruments, Agere Systems and Level One / Intel. His experience in networking and communications includes a variety of technology areas including digital signal processors, network processors, modem chipsets and DSL transceivers. Mr. Garza holds a Bachelor of Science degree in Electrical Engineering from the University of Texas at Austin. He has also served on the Board of Directors of the Network Processing Forum and participated in DSL standards work at the International Telecommunications Union (ITU).