Large investments in SDI core routers and other edge devices serve a legitimate function, writes Will Waters, Senior Director for IP Workflow Strategy at NewTek, as he maps out a viable hybrid path towards IP-based workflows.
It is an exciting time for broadcasters. Around the globe, broadcasters and video production professionals are navigating the transition to IP, perhaps the biggest change the video production industry has seen in the last half-century. Where once a tiny number of producers generated video, now almost anyone can make a show and broadly distribute it across IP networks and the internet. As a consequence, consumers of video now expect more content, in more places, and they want it on their own schedules.
To be successful, broadcasters today must be able to quickly and easily produce content that is curated for many smaller groups of viewers, and do so while reducing costs. After all, the true end goal of the transition to IP in broadcast is to gain the flexibility, efficiency and cost savings that the larger IT industry can provide.
“Since the reality of the transition to IP is hybrid for the foreseeable future, this means there is a real economic reason to consider inexpensive and widely available network infrastructure”
Recently, the ratification of major standards such as SMPTE 2110 has begun to paint a picture of how IP solutions will emerge. What this means for broadcasters is that video production over IP is no longer a future milestone, but a reality today. Other areas of the modern production facility have long been using networked IP infrastructure in their workflow. Software-driven applications within the live production space, like graphics generators and video playout servers, are already connected to the rest of the facility via the network for interface and control, leaving only the raw audio and video streams in an SDI format. The ratification and adoption of IP standards for moving high-quality, low-latency video suitable for live switching now brings the rest of the studio into the same fabric.
Supporting SDI-based gear
However, the need to support existing SDI-based gear is clearly apparent for a successful transition to IP, for several reasons. Large investments in SDI core routers and many edge devices serve a legitimate function, and there is little reason to get rid of them quickly. The HD signal types commonly used are HD-SDI (1.5Gb/s) and 3G-SDI (2.97Gb/s in progressive format), and these uncompressed formats can be moved into 10Gb/s network routing easily. Higher resolution formats such as UHD1 (12 Gb/s) and UHD-2 (24Gb/s) require a different approach and are the topic of a different debate.
Further complicating matters are various uncompressed IP formats introduced prior to the ratification of SMPTE 2110 and the introduction of 12G-SDI. As a result, some manufacturers are resisting putting IP-capable I/O on their equipment. This dictates that a proper transition to IP will require hybrid systems taking advantage of the best-of-breed devices, whether they are SDI- or IP-based.
SDI conversion to and from IP is an important design consideration for engineers as they plan out the IP infrastructure. While it is a practical reality that conversion will be required in a hybrid facility, the amount of conversion should be weighed against the cost of full infrastructure switchover. Depending on the type and capabilities of the conversion products, there are hidden costs when dealing with multiple resolutions and formats. Additionally, lots of conversion usually comes with increased power consumption, adding to cost.
IP deployment impacts workflow
With IP, all traffic in a facility can be on a single common infrastructure. However, transitioning to IP for the sake of using IP alone gains little benefit in both workflow and investment. If IP implementation does no more than provide a simple, unidirectional connection between devices, perhaps reducing expense, it may not justify the effort.
Once studios are connected on the IP network, however, the normal physical restrictions that traditionally impact storytelling and content production go away. The studio itself is no longer limited to the distance between the video mixer, graphics and media playback systems, recording devices, camera controllers, audio desks and other necessary gear.
Likewise, video routing for monitoring outside of the studio is much easier. If someone on one side of the building desires to view a source on the network, they can easily do so without touching a single cable. When each output required discrete physical hardware, this was an unavoidable requirement of system design and manufacture. For IP-connected devices, however, the number of channels a system handles is freed from such physical constraints. If there is no need to view a source, it does not have to transmit across the network.
Consider a typical graphics system with 100 pages of prepared graphics ready to display. All 100 graphics pages run in parallel over IP, allowing users who wish to freely display any one or more of them at any time. Critics may point out that this example ignores a key fact, that even the most powerful current graphics systems are unlikely to be able to render 100 simultaneous channels of output at one time. However, this is where design considerations in the IP world come into play. A very important aspect of the IP world is the bi-directional connection between devices. Going back to our graphics system example, the source can be alerted, effectively instantly, that a request has been made for connection to one of the 100 channels.
Let’s consider the benefits of SMPTE 2110. SMPTE 2110 incorporates the concept of ‘essences’, where elementary streams for synchronisation, audio, video, metadata and control are all available to connected devices and can work with only the necessary part of the stream required. Moving uncompressed video, audio and data across the facility for specific needs is possible. Not only is this interesting to engineers from a cost-saving standpoint, compressed formats are required for fully virtualised software-driven production systems.
Once signals are in the proper format, other design considerations exist. What kind and type of network should engineers invest in? This is yet another area of debate. If uncompressed video is required, then 10/40GbE or even 25/50/100GbE connectivity is necessary. Fully redundant systems double the infrastructure. The data bandwidth and networking protocols required for uncompressed transfer also usually involve separate logical networks for the elementary streams and timing protocol. These are not detrimental requirements; however, the heavy up-front investment arguably defies the market pressure of creating more content and delivering to more places with no more cost.
Fortunately, by using networks already in place or reasonably available to purchase, live production workflows based on IP can be realised to train staff and refine business models. Investing today does not limit the decisions in the future. The advantage of the IP studio is that the infrastructure is agnostic to the protocols that travel across it. From a network perspective, it becomes a matter of adding bandwidth and processing to reach the needs of the protocols on top of it. This is fundamentally different for the broadcast industry, as previous advances in resolution and format required wholesale infrastructure change.
“The HD signal types commonly used are HD-SDI (1.5Gb/s) and 3G-SDI (2.97Gb/s in progressive format), and these uncompressed formats can be moved into 10Gb/s network routing easily”
Since the reality of the transition to IP is hybrid for the foreseeable future, there is a real economic reason to consider inexpensive and widely available network infrastructure. Operators may consider using gear that can operate on IP networks for new or smaller live production environments that can then be attached to the traditional facility infrastructure with conversion. This allows operators and engineers to experiment with IP as the new signal types, and lets producers create new and exciting shows with innovative workflows previously unavailable.
Live production is a real-time art form, spontaneous by its very nature. The more availability to sources and content, the greater the creative options. The transition to IP is no longer a concept based in the future, but rather one that can be built on today, which is why it is an exciting time for broadcasters.