Migrating to IP infrastructures challenges every media professional.
IP has almost been positioned as a digital nirvana for both content generation and transport. We’ve been told IP is cheaper, faster, better and… add your own descriptor. While that eventually may be the case, getting there remains a challenge.
The conversion to IP-centric facilities and transport systems was again a key theme at this year’s NAB trade show. Hundreds of companies touted their IP equipment, solutions and expertise.
But, when migrating to an IP facility what are key factors to consider? Does the move to IP need to be a forklift upgrade or even a total greenfield build? Can I migrate my older equipment into a newer IP infrastructure?
What differentiates IP solutions? Is an IP switch from one vendor the same as one from another? What features should I look for? Finally, many of today’s proposed solutions are promoted as using COTS (Commercial -Off-The-Shelf) equipment. Is that good—or bad?
To help readers better understand some of the key issues surrounding IP migration, The Broadcast Bridge asked John Dale III, newly appointed CEO and President of Media Links about the migration to IP networked infrastructures. Because Media Links has deep knowledge in IP transport and interface technology, his perspective may be useful as you plan your facility’s evolution to an all-IP world.
The Broadcast Bridge: What are some of the most basic issues facing technology managers as they plan their migration to IP-centric transport solutions?
John Dale: The current industry practice of relying on baseband SDI interfaces for access to transport networks is outdated. Video production, playout and transport is increasingly moving to IP-centric solutions. For instance, many of the new OB trucks are already highly IP centric. These production centers work within the IP domain and can easily output standards-based IP signals.
An IP demonstration project conducted by Belgian broadcaster VRT and the EBU created an in-house R&D testing lab called VRT Sandbox. The interoperability experiment was designed to prove that audio and video production workflows based on IP and leveraged by applicable standards can be more efficient than a comparable SDI-based infrastructure.
One discovery was that, as of now, a Live IP production environment currently is about 7 percent more expensive to build than a similar SDI plant. This is primarily due to the need for hybrid equipment. Even so, the long-term benefits of IP far outweigh the extra cost, which will certainly be reduced as IP becomes mainstream.
But not everyone can build a “sandbox” to try out new technology. And, we can’t simply throw everything away. Engineers and managers need a transition path that allows existing equipment to be used until it reaches end-of-life, while incorporating newer solutions for the duration. They want a smooth migration to the IP future.
The Broadcast Bridge: Some vendor’s solutions look more appropriate for greenfield sites, but most facilities will not have that option. Can one migrate to IP without a forklift renovation?
John Dale: A greenfield solution would be great, but it’s not required. The key to a smooth transition is to leverage the greater flexibility (and lower prices in many cases) of IP backbones and gateways allowing hybrid operation (SDI and IP). But you need a proper migration plan.
The transition from SDI to IP can be broken down into three overlapping components.
First, migrate from traditional transport SONET/SDH connections to IP-based connections. Second, include a flexible gateway which can accept and process both SDI and IP media. And finally, transition from traditional IP routers and switches to SDN (Software Defined Networking) or flow-based technology.
The Broadcast Bridge: When it comes to moving media, what needs to be incorporated in the solutions? How can a technical manager plan for the multiplicity of future formats like 4K/UHD, HDR, HFR and OTT?
John Dale: A comprehensive transport networking solution needs to support multiple types of interfaces and features. It should support baseband video and audio signals, from standard definition up to UHD and include both compressed and uncompressed media, but also be capable of accepting the IP/RTP version of these signals. Ultimately, the platform will need to support conversions between multiple signal types, acting as a media converter or in some cases as an encapsulation converter. So, the solution is not just a fixed configuration but includes virtual capability to update as needs change in the future.
Also, don’t forget about managed and guaranteed data services. These day’s significant amounts of supplemental media file content and other data is needed as part of the production and to accompany real time video/audio transmission. Feedback we receive continually highlights the growing need to transport more and more data in addition to the myriad broadcast video/audio formats, resolutions and compression methods we support.
Media Links has a white paper, “Media Network Migration to IP-Based Transport,” which provides a rich background in the issues and solutions for the IP transition. It can be downloaded here.
The Broadcast Bridge: This year’s NAB trade show was filled with exhibitors touting their IP solutions, many based on COTS routers and switchers. First, are COTS routers and switchers capable of reliable high-bandwidth signal transport? And second, are all COTS platforms pretty much equal?
John Dale: This area continues to confuse end users. It is possible to use COTS routers and switchers for media, however there are some very important considerations. Each COTS switch or router has been designed for a specific or multiple specific applications. Generally these applications are IT based and not Media specific.
Also, these applications may not have the sensitivity to flow management, blocking, forwarding rates, packet loss and system timing that real time media applications need. When used for media applications, extra precautions and limits may be required. Per flow management, lower throughput to match forwarding rates, and memory/queue management to avoid packet loss are just some of the additional considerations.
Therefore, much of the gear using COTS technology may be insufficient to meet the demands of media switching and routing. These devices simply do not have the ability to properly transport and manage continuous multi-gigabit streams while maintaining adequate high levels of QoS and reliability.
In fact, high-speed Gigabit multicast streams, which are common in production and broadcast spaces, represent a particular challenge for even some enterprise-level IP routers and switches.
Figure 1. This block diagram illustrates how a high-bandwidth IP connection enables a remote broadcast to be handled from the studio location. Such a configuration can greatly reduce costs by keeping most of the production team at the home studios. Click to enlarge.
The Broadcast Bridge: What about uncompressed signals? How should they be handled?
John Dale: In transport, there are two main technical concerns; quality and latency. Every content originator would prefer that their content be carried with minimal or no processing. Hence the phrase “uncompressed, bit perfect, absolutely minimal delay” transport of the media. However, depending on end points and distance, there can be significant expense handling the bandwidth of media, especially 1080p and UHD-1 & -2 signals.
In general there are 3 modes:
- As fully uncompressed from end-to-end utilizing the most bandwidth.
- Lightweight, very low latency coding, which is a good balance between bandwidth and performance/quality.
- Heavier coding with sub-to-multiple frame latency. Good trade off when full bandwidth is not available.
Each of these three modes can be used at different times during the SDI-to-IP migration period. Combined, they provide sufficient flexibility to accommodate multiple signal formats at different times within the transition period.
The Broadcast Bridge: What about packet processing and conversion?
John Dale: Packet processing is possibly the most critical aspect of the transition to a fully IP end-to-end network. For security, and the support of different addressing schemes and overall traffic management, individual IP packets may need to be converted as they pass from endpoints into the backbone or from one network domain into another.
This typically happens when packets pass from a closed network, like a production studio, into a backbone interconnection link. As the network evolves, the types of transformations may change, but some form of adaptation likely will be required throughout the life of the network.
Further complexity may be added if the encapsulation method is not consistent from end-to-end. For instance, a service might start out as SMPTE 2022-6 and terminate as SMPTE 2110.
The Broadcast Bridge: What functions does the equipment interface perform?
John Dale: Some of the major functions performed by the interface equipment include:
- Conversion between different network numbering plans, Network Address Translation (NAT).
The process might be as simple as converting IP addresses and port numbers. Or it could be more involved such as converting between IPv4 and IPv6 networks. This feature is useful when connecting between groups that may only occasionally share content, such as a pooled news feed being distributed to multiple broadcasters.
- Providing intelligent management of Ethernet/MAC addresses and VLAN tags to support stream prioritization and flow management.
- Managing unicast, multiple unicast and multicast connections.
- Managing assurance and logging at each demarcation point. In the event of an error in the workflow, the operations team needs to determine the source and correct it as quickly as possible. The gateway needs to provide error status on streams in both directions, ingress and egress.
- The interface needs to be flow based, which enables intelligent forwarding and filtering of packets based on more than simple IP addresses or VLANs. To achieve compatibility with SDN, open APIs (Application Programming Interfaces) can provide the granularity of control required by this powerful new tool.
Figure 2. A switched IP network can support multiple remote broadcast locations, all fed back to a central studio. The benefits include the ability to utilize switched backbones and on-demand network bandwidth. Click to enlarge.
The Broadcast Bridge: What do future networks look like?
John Dale: Broadcasters, production centers, and all media facilities will continue to move to IP-based facilities. It is only the quickness of this change that remains unknown. However, media transport networks will need to evolve even faster in order to handle the new media content producers will generate.
Instead of just baseband SDI and HD-SDI video interfaces, transport networks should install native IP-video interfaces, which support both compressed and uncompressed signals. There will still be the demand to transport baseband video and audio so the network will be a SDI/IP hybrid for the transition period, which may last multiple years.
Transport networks will need to support uncompressed encapsulation as well as lightweight coding like JPEG XS as well as heavier coding like JPEG 2000 (VSF TR-01) and H.264. The range of incoming audio signals will likely be varied as well. AES3 or AES10 signals today will be encapsulated into RFC 3190 or AES67 in an IP network.
The network may need to support PTP (Precision Time Protocol) in order to keep video and audio signals synchronized. Today frame synchronizers are used to sync wild feeds back at the broadcast center, but with the expansion of IP connectivity, PTP can be used to time stamp packets so that they can be easily aligned back at the studio.
And again, don’t forget about low latency guaranteed data delivery, especially protecting it via hitless switching as practiced with high value video/audio content. Modern networks will also have to support QoS and QoE signal monitoring.
The Broadcast Bridge: What does your crystal ball predict about the industry’s migration to IP networks?
John Dale: The move to IP-centric networks and production equipment is inevitable. With that change comes new flexibility. It is possible to effectively transition from today’s SDI networks to IP-native architectures that rely on SDN, per flow controlled/switched backbones.
This is not to say the transition to IP will be easy or straight forward for everyone. Closed or private networks, like those used in production and playout environments will be easier to convert to IP than WAN environments, which are more complex.
Look for solutions that comprehensively incorporate IP connectivity throughout their internal structure and external interfaces as well as support baseband interfaces. Be sure interfaces provide the ability to seamlessly transport non-IP signals in a standards-compliant manner. Look for the ability to add new features and functionality over time.
By selecting such equipment, you will enable the migration from SDI to IP to be completed as fast, or as slow, as needed. In addition, a carefully-developed plan will ensure you can still support existing applications and clients while protecting flexibility to adopt future standards, interfaces and protocols.
Figure 3. The Media Links MD8000 illustrates a modular solution that provides a flexible system architecture making it easy to change signal and network interface types. Click to enlarge.
The Broadcast Bridge: What is your bottom line advice for technology managers facing the migration to IP? Is it time to panic, pause or proceed?
John Dale: Facility managers need a plan, particularly in the selection of a backbone network technology and in choosing the correct equipment to implement the service interfaces and stream switching. They also need to consider vendors with experience along with a solid track record and reputation in broadcast IP, not just legacy video or data transport.
Finally, these transition components should be considered:
Develop a plan to incrementally move from SDI video to native IP video streams. Replace older SDI technology with newer network endpoints equipped with SDI/HD-SDI interfaces that packetize incoming streams. This makes them compatible with the IP backbone, using a standard such as SMPTE 2022-6 or 2022-2 and with a future towards 2110.
At the destinations, convert back to SDI although sometimes the customer may request the media be output directly as IP-wrapped video. Going forward, more IP-native signals will be supplied for transport. As demand decreases, the SDI interfaces can be migrated to other locations that still require SDI-based services, and then eventually retired.
You might also like...
Outside Broadcast connectivity using managed and unmanaged networks is delivering opportunities for employers that enhances productivity through flexibility, scalability, and resilience.
Founded in 2001, Blackmagic Design is based in Melbourne, Australia, with offices in Fremont and Burbank, Calif. in the U.S., as well as Manchester, England, and Amsterdam, The Netherlands.
We continue our series of articles on design considerations for Media Supply Chains looking at the technology trends influencing Ingest.
Signal transducers such as cameras, displays, microphones and loudspeakers handle information, ideally converting it from one form to another, but practically losing some. Information theory can be used to analyze such devices.
Connecting a camera in an SDI infrastructure is easy. Just connect the camera output to the monitor input, and all being well, a picture will appear. The story is very different in the IP domain.