Remote contribution has been a mainstay of broadcasting since the early days of television, and getting the contributed material back to the production data center in a timely manner is still a technical challenge.

The Broadcast Bridge’s Live Sports Production Themed Content Collection has already examined the background of remote contribution; this time, we will focus on standards.

Historically, analog video was transported over a leased line. Modern solutions add IP connectivity and digital streams which provide more flexibility. Closer integration between the remote location and the production data center is possible with shared metadata, especially when cloud technology is part of the mix. This facilitates better management of the bulkier essence (media).

The functionality of the key network hardware devices (hubs, switches and routers) facilitates the design. A variety of network topologies are optimal for different applications and choosing the wrong device or configuring the wrong topology compromises the entire contribution process.

Therefore, understanding network theory is key foundation knowledge for broadcast engineers developing IP based studio and contribution solutions.

Three Common Models For Remote Contribution

There are many variations on the theme of remote contribution but we are going to explore three common models as well as flexible options for processing compute. The best option depends on the duration of the outside broadcast event and the number of different simultaneous feeds required.

All three models are affected by latency, which is a natural consequence of using digital technologies. It was always there in the days of analog TV but the delays were mainly introduced at the switching matrices and only added a frame of delay here and there.

Digital systems introduce much more latency when moving content, compressing it, processing it and packaging streams for deployment.

Choosing the wrong solution at any point can require buffering later on to correct synchronization problems. This adds significantly to the delay.

Models 1A & B - OB Trucks

A broadcaster might deploy a truck (or fleet of vehicles) fitted with production facilities, control rooms and multiple camera support. These are driven to the event location for deployment. This is appropriate for a major sporting event, rock concert or festival scenario where there is no pre-existing embedded video facility. Everything needs to be taken to the location and connected together onsite.

There are two ways vehicles are commonly utilized and both scenarios support large-scale production, better quality and higher resolution.

Scenario A is to send all of the crew with the vehicle and produce locally in the traditional manner.  The concentration of processing resource and crew locally reduces the amount of connectivity required between the venue and production center because produced feeds are presented for backhaul.

Scenario B is what is often referred to as a REMCO model where all of the processing tools are located in the vehicle on site but much of it is controlled remotely from a remote team in a production center. Camera operators and some of the crew will be on site. This also reduces the amount of connectivity required because remote control data requires less bandwidth than media streams and once again only produced feeds are presented for backhaul.

Provided sufficient bandwidth is available, these scenarios would work well for events such as the Olympics where multiple simultaneous sports happen. Live sports production and major events such as the Olympics always test the extremes of what is possible, and this is why major innovative steps forward in technology development happen every time the Olympic games are televised.

Events on this scale often benefit from a locally deployed production resource to reduce the content sent back to base. Preparations for the Olympics begin years in advance and are usually organized using a single host production hub which then syndicates feeds to broadcasters globally. Scalability is possible by integrating multiple OB units via that central hub.

Everything being filmed can be recorded locally for future use but not all of that needs to be contributed for broadcast right away.

Model 2 - Hybrid Remote Deployment

Operationally, this requires less equipment to be deployed. Some crew would still be onsite and shooting with operated cameras that are physically or wirelessly connected back to a specialist capture OB vehicle. This vehicle sends the content back to a production data center as individual essence feeds via a fast Internet uplink, possibly to a private cloud. The onsite crew are then virtually connected to the same workflow space as the data center production team. This improves the level of collaboration.

Hybrid operation can be a lower cost solution than deploying a fleet of full OB vehicles. Very small OB hubs can be constructed in rapid deployment vehicles, and modern technology can integrate more cameras, IP networking and local control room support into a small van.

This approach requires higher bandwidth and low latency connectivity between venue and remote production center.

Model 3 - Full Remote Operation

This is the lowest cost solution although the initial outlay for installing the onsite systems may be quite high.

Some venues have embedded video production systems already installed and running under their own management that deliver an outgoing feed when frequent televised events take place. A compatible feed can be routed directly to the broadcast network, and this model is most appropriate for equipping a stadium with embedded video facilities. It is most useful when there are regular televised events from the same location.

Multiple cameras would be mounted on remotely steered gimbals controlled from the remote production center. The cameras can be installed with excellent fields of view in locations that are quite inaccessible for a human operator. Other cameras can be flown on wires to bring them down to the center of the arena for close-up shots. Some operators may be physically present for camera operation especially with roving hand-held wireless cameras.

The effects of latency need to be managed especially carefully. Not only are there challenges with getting the content back to the production data center, but remote commands to operate and steer the cameras must be delivered with as near-zero latency as possible or they cannot be operated effectively.

Compute Resource Models

With most of the approaches we have just described it is common practice to locate all of the heavy lifting compute processing resource for functions like switching and replay in one location; either the truck, or a remote production center (with the potential for the remote production center to use either on-prem, owned off-site or public cloud compute resource). Control over the compute resource may be remote given sufficient bandwidth and low enough latency of connectivity.

The reason for this is that it reduces the challenges of managing latency. Deploying multiple resources over multiple network paths may cause traffic to arrive out of sequence due to some network routes being longer than others. However, with the ongoing rapid advances of connectivity and data center compute technology that approach may change.

Public Cloud Options

A production content management system stores assets in a shared infrastructure. The content is then accessed from anywhere it is needed – potentially facilitating elements of the production crew to be at multiple locations.

Using the public cloud as the location for this content management function may be a viable option in some scenarios. Using the public cloud as the location of production applications and associated compute processing resource may also be viable in some scenarios.

One of the potential benefits, possibly for smaller and more nimble productions, is that it potentially avoids the need to have a dedicated data link from the remote location all the way back to the production center. The data link only needs to reach a suitably robust Internet access point for successful ground-to-cloud contribution. Similarly production operators only need suitably robust internet access for proxy monitoring and control data.

Backhaul

The term backhauling describes a high-capacity hard-wired link between the remote site and the production data center. This needs to be set up well in advance of an event. The available capacity must be sufficient to carry all the required media streams and metadata, and this too can be used to manage latency as a higher bandwidth link supports lower compression ratios which should introduce less latency.

Backhaul circuits utilize wholesale commercial access to an internet exchange point. Decide beforehand what capacity and latency is required for the circuit and book the service well in advance.

Network links are not all created equal. A bottleneck might still occur with multiple links if they are aggregated together inside the telecoms provider’s network. Genuine redundancy and triangulation are more likely if different providers are employed.

Uploading speeds are often significantly slower than downloading speeds, and high throughput uploads are necessary for contribution. This might rule out some alternative options right away.

Here is a summary of the different kinds of backhaul delivery medium describing their relative performance. These specifications are representative but differ according to which source you consult:

* Data rates for 4G, 5G and private 5G vary and are described in this article.

The optimum choice for the backhaul point-to-point connection is clearly an optical fiber network if it is available.

Multiple redundant circuits can be aggregated to provide greater capacity. This is more robust in case any of the circuits goes down, but multiple circuits may cause traffic to arrive out of sequence due to some network routes being longer than others. Receive buffering reassembles packets into the correct order but increases the delivery latency.

In part two in this series we explore how network design plays a big part in network efficiency, and how the configuration of hardware devices such as hubs, switches and routers makes a huge difference. With a variety of flexible topologies capable of moving data around the network, understanding network theory is key.