Hot Button Discussion
SMPTE ST 2110: IP Revolution's Next Step
By Michael Goldman
The quest to make possible the wide-ranging adoption of IP-based professional production networks has greatly accelerated in recent months, thanks to concerted standards work across the industry. That work has led to SMPTE’s recent approval and publication of SMPTE ST 2110, suite of standards for Professional Media Over Managed IP Networks. According to Paul Briscoe, principal consultant at Televisionary Consulting, the arrival of ST 2110 makes possible wide interoperability of IP-related tools from vendors across the spectrum using a common data infrastructure and permitting the separate transmission of video, audio, and metadata as needed across those networks. Briscoe is a SMPTE Fellow, SMPTE Canadian Regional Governor, chair of the SMPTE 24TB-01 Lipsync ad hoc group, and a member of the ST 2110 Working Group.
“SDI digital has served us through the evolution of HDTV, and is still serving us for UHDTV,” Briscoe says. “But recent technology evolutions have reached the point where IP networks are now suitable for the kind of bandwidths we need to build similar scale systems as we do with SDI, to accommodate higher bit rates that UHDTV brings, and offers us virtually unlimited scalability and size through the technologies and techniques available in the IP domain. Therefore, we can leverage everything that is going on as a result of [the IP revolution and technology development] and not have to re-invent the wheel. Instead, we can take our unique value-add, which involves an intimate understanding of how to handle media essence, how to convey it and synchronize it, and build on top of that foundation.”
“ST 2110 allows us to replicate the existing systems we build today with SDI—
to emulate them entirely. Essentially, we can take an SDI signal, fully transport it as independent essence streams over IP, and put it back together anywhere else and make another SDI.”
This, Briscoe emphasizes, makes ST 2110 potentially superior over the existing SDI over IP standard, ST 2022-6, as far as content creators are concerned. “That’s because, ST 2022-6 has to transport every single bit of SDI, including all of the timing and blanking,” he explains. “Whereas, ST 2110 takes the video/audio essence apart, we can transport video on one stream, audio on another stream, and metadata on yet other streams. That opens the door to [many advantages] such as taking audios and sending them off independently into an audio sub-system, without the burden of all the video overhead of SDI; or taking a closed-captioning stream and sending it to a service in the Cloud over IP, since IP connects us to the Cloud now, and that cloud service could translate it to many languages—both text, and spoken word—and send the resulting caption and audio streams back for multi-language program integration into the system workflow.”
In other words, Briscoe states that with ST 2110, “we can now build highly efficient and flexible media systems, which move around and deal with only the essential pieces needed.”
ST 2110 has many parts, he explains. They include ST 2110-10, which describes the overall system—“an umbrella document,” Briscoe’s states. ST 2110-20 describes the transport of full raster video, meaning that “we are taking the entire active picture and none of the ancillary data space—we are just transporting the pixels.” And then, ST 2110-30 references the transport of audio over IP, utilizing the AES67 specification, “but a special, slightly constrained version of AES67 to better accommodate TV purposes.” The next core piece is ST 2110-40, which transports ancillary data over IP. The final core document, ST 2110-21, is a timing model for video.
“Because video data is so dense, a relentless torrent of data, it was deemed appropriate to develop a model that describes the transmitter’s timing behavior, and implies a receiver buffer model,” he elaborates.
At press-time, SMPTE has published the -10, -20, -21 and -30 documents. The -40 document is presently undergoing final committee draft ballot and is anticipated to be published early this year. “With the publication of 2110, we will see a lot of  products at NAB, no question,” Briscoe says. SMPTE is posting the current status at smpte.org/st-2110.
Briscoe states that ST 2110’s entire development and approval process “was done in record time—less than two years. That was based on having a solid proposal from [an original Video Services Forum document called TR-03], which was submitted to SMPTE for standardization. That document contained the core of what ultimately became the ST 2110 suite.”
The second reason why the ST 2110 approval process was able to work so efficiently is that, “we didn’t invent anything anywhere we didn’t need to, and we focused our energies on our own particular SMPTE value-add,” Briscoe says. “That allowed us to go faster than a standard suite of this complexity would normally have gone. Thus, we didn’t spend time arguing over whether we should, for instance, invent a new video transport method. One existed in the internet world, so we decided to leverage that method, based on the well-articulated VSF proposal.”
Along those lines, one of the key elements of “not re-inventing the wheel” involved strategically packing ST 2110 with existing standards such as the AES67 audio transport specification, and another major component within the broader ST 2110 umbrella—the addition of SMPTE ST 2059. That standard describes how to synchronize video equipment over an IP network, built on the foundation of the IEEE 1588 Precision Time Protocol (PTP) specification.
“2059 is a published standard for over two years now,” Briscoe explains. “In fact, although it is entering its third year, we are just now completing the one-year review activity. That’s because we were delaying that process while we did inter-op testing of 2059 in concert with 2110.”
ST 2059 was published in two parts—the first defining signal generation based on time information delivered by the IEEE 1588, and the second part defining an operating profile for PTP optimized for timing and synchronization of professional media streams. According to Briscoe, this development proved crucial to the overall success of ST 2110, even though it was put together long before anyone knew how the development would play out.
“After several years of research by the SMPTE/EBU Task Force, we realized that we could use PTP as the foundation on which to build our media-specific reference virtualization, once again using existing and proven infrastructure,” he says. “Pure PTP from the IEEE has a profile provision, which allows a set of parameters that define the operation of PTP and allow the transport of user-specific metadata. There are different profiles for different industries, so we created one for broadcasting called the SMPTE Profile [defined in ST 2059-2]. It describes the behavior of PTP that we want on a broadcast network—how often messages are transmitted, how communications are established, and so on. Plus, we also transmit a bit of metadata in this profile, which handles certain elements that are not inherently built into PTP. For example, although it seems obvious when you hear it, we had to include the default frame rate of the system. Typically, you just pick up a genlock signal, and any piece of equipment can discover this, but with PTP, there is no way of knowing what TV system you are using. A master generator has to tell a slave through a message in PTP. We pre-tuned PTP in ST 2059 for our industry, and that same work plugs directly into ST 2110, allowing PTP to become the replacement for genlock reference in an IP-based system.“
“Another important component of ST 2059 is that it not only provides reference signals virtually in an IP system, just as they would be provided in an analog system, but it can also be used directly with analog signals still in use at existing facilities. Thus, the transition to PTP can be seamless for the broadcaster—their new PTP master synch generator for the IP side of their facility will still produce all the legacy signals needed. It will also produce the SMPTE [Profile] PTP needed by the IP side of the system, and they will be harmonized, synchronous, and in concert with each other. From a system timing point of view, it all looks like one holistic system.”
Briscoe adds that there is another large slice of the overall ST 2110/IP pie. That piece is, essentially, “the software side of 2110, if you want to call it that,” he says. He’s referring to work currently underway by the Advanced Media Workflow Association (AMWA) called Network Media Open Specifications (NMOS), which is designed to increase control plane interoperability.
“NMOS consists of three pieces,” he says. “One is called IS-04, which is the registration and discovery protocol, so that every ST 2110 device on a network, when you plug it in, will talk to a server and disclose its capabilities. The second piece, currently underway, is called IS-05. That piece translates a request for connecting a source to a destination to the necessary network transactions required to cause that connection to occur. It’s essentially the interface for connection management—what your routing switcher control panels talk to. And a third piece, under development, is called IS-06. Its purpose is to define a common API for Software Defined Networking (SDN). Currently, routing inside the switch is done through normal multicast routing techniques. But with SDN, you can gain greater switch utilization, manage latencies, and accrue many benefits by tuning the routing according to what you know about what is going on. So, IS-06 will be another interface, an API to a software-defined network controller.”
Briscoe adds that these three pieces, combined, “will allow ST 2110 to be built into an intelligent system. By itself, ST 2110 is just a set of protocols, but these three pieces from AMWA will be the intelligence side of the system, within the greater fabric. The idea is that, using these NMOS tools, vendor control systems can now connect to the network all in the same manner through a method that allows them to collaborate. Right now, to achieve control interoperability, if you have a Brand X router and want to use Brand Y control panels, a protocol gateway is needed. This system provides one single way for vendors to control the routing. That broadens interoperability hugely, and enables best-in-class customer equipment choices.”
These developments “are not the end of the story” for ST 2110, as “additional work is already on the table” in various additional categories, he says. He also believes that ST 2110’s ongoing development and rollout process stands as something of a potential template for modern standardization work related to IT-based workflow standards.
“ is probably among the most complex set of documents that SMPTE has ever done,” he says. “This is deep, network stuff—a living, breathing network. My strong recommendation [for future standards work] is to put the egos aside and look outside first. We’ve always developed our standards almost entirely ourselves, out of necessity, but today we live in a rich technological ecosystem. If we can use or adapt something that exists, then we should do that. Our value-add is not the entirety of our final product, it is the specific capabilities, usefulness, quality and ultimate value to the customer of the final product. By focusing our energies on things we know the best, we can go further and deeper than if we did it all ourselves. And there are bonuses to using existing technologies. We can use technologies that are already more tested than our standards will be across their entire lifetimes—think about IP. The ubiquity of IP and its ecosystem makes it both robust and cost-effective. ST 2110 should serve as a good model for the efficient development of standards.”
To aid in the understanding of ST 2110, SMPTE has introduced a new virtual classroom course “Understanding SMPTE ST 2110: Live Production of Professional Media Over Managed IP Networks,” an educational program focused on SMPTE ST 2110. This SMPTE Virtual Classroom course will provide attendees with the knowledge they need to help their companies make the most of SMPTE ST 2110. The new course assumes a basic understanding of IP and associated terminology. The SMPTE Virtual Courses “Introduction to Networks,” “Routing and Switching Essentials,” and the “Essentials of IP Media Transport for Broadcasters: Moving Real-Time Video and Audio over Packet Networks” are not required to register for “Understanding SMPTE ST 2110.” However, these courses enable a deeper understanding of IP, internetworking, and many essential concepts that are referenced in the ST 2110 suite of standards and are recommended prior to enrolling in “Understanding ST 2110.” Registration, details, and the full schedule of SMPTE Virtual Classroom courses for 2018, including SMPTE’s “Introduction to Networks,” “Routing and Switching Essentials,” “Essentials of IP Media Transport for Broadcasters,” and the new “Understanding SMPTE ST 2110: Live Production of Professional Media Over Managed IP Networks,” are available at smpte.org/courses.
AIMS and MNA Merge
Members of the Alliance for IP Media Solutions (AIMS) and the Media Networking Alliance (MNA) voted in mid-December to formally approve a merger between the two organizations. As explained by TV Technology, the move comes after the two groups spent about two years working closely to promote a standards-based approach to interoperability, an effort that helped get the SMPTE ST 2110 standard approved. According to an AIMS press release, the merged organization will keep the AIMS name and bylaws, but member companies that belong to both groups will now only pay a single membership fee. Officials for both organizations say the merger will allow a larger combined voice and a single unified destination for organizations, companies, and individuals to participate in the pursuit of an IP-based industry future.
Net Neutrality Axed
The politically controversial drive to bring an end to so-called net neutrality rules governing how ISP’s conduct business in providing Internet access to consumers came to a head in mid-December when the Federal Communications Commission (FCC) formally rolled back those regulations on a party-line 3-2 vote. According to a Broadcasting & Cable analysis, in addition to revoking those rules, the FCC voted to reclassify ISPs, wired and wireless, as non-common carriers, thus exempt from additional existing regulations, and left most internet policing functions up to the Federal Trade Commission or Justice Department. Further, the move includes the FCC’s declaration that it has the authority to over-rule state or local net neutrality regulations. ISP providers were pleased with the ruling, but the article suggests they are expressing a willingness to work with Congress to legislatively address some of the concerns the new approach is causing about the possibility of a non-democratic internet. Indeed, Studio Daily’s Bryant Frazer recently offered up an analysis of some of the entities that may potentially be hurt by the ruling.
Small Drone Rule Returns
In the latest chapter of an ongoing saga, the Federal Aviation Agency (FAA) recently reasserted its right to require owners of small drones to register with the government any flying device that weighs more than 0.55-lbs. It was a rule many tech companies favored, but pushback from consumers and hobbyists ended up bringing the rule to court. As Verge recently reported, a Federal Appeals Court struck down the registration rule in May 2017. However, in early December, Congress made the rule a law by adding it to the larger National Defense Authorization Act, which gives the FAA broad powers, including jurisdiction over small, unmanned aircraft.