Following the great migration to remote production

By John Dale, director and chief marketing officer, Media Links

During the early days of the pandemic live sporting events were basically shut down. Now, we are seeing an increase in live events with limited on-site spectator attendance and experiencing renewed active planning for the Summer Olympics, which were delayed by a year and are now scheduled for 2021.

We absolutely see an increase of utilisation in remote production, which can mean a lot of different things to a lot of different people. In general, it’s more feeds; more individual camera feeds, more data and more bandwidth as a result. There is less talent and less personnel on site, resulting in more feeds in both video and data connectivity. We are witnessing this where broadcasters are using their own services or using a wide-area service provider to supply JPEG-2000 services.

There is an increased use of 1080P, especially in the Americas, where we saw very little of that in the past, as well as more use of 1080P potentially due to its clean up-conversion to 4K. We’re also seeing multiples of 1G or greater than 1G data connectivity to go along with the video connectivity. This migration to remote production is really pushing the need for lower latency, not just in coding latency but in route miles in terms of taking the shorter distance and more express/quicker fibre paths.

Overall, we see a significant increase in bandwidth to support the transition to remote production. This favours 100Gbps connectivity to the event locations to address the need for more camera feeds and more data feeds; 100Gbps capability replaces multiple 10Gbps capability. Some of the venues right now are configured with dual 10Gbps – primary and protect capabilities – so these remote production applications are definitely exceeding that. They may not get up to 100Gbps, but they are somewhere between 10Gbps and 100Gbps. Forty kilometre and 80 kilometre 100G optics are being introduced and certainly it is now possible to achieve 100Gbps connectivity over essentially the same fibre links that were running 10Gbps.

Latency and codec choices

We’ve seen the use of JPEG-2000 quite substantially for backhauls of video from premier events. JPEG-2000 at the appropriate bandwidth is really good quality and has been referred to as the quality benchmark. However, there is some latency with JPEG-2000 in the order of magnitude of about 100 milliseconds. For lower latency applications JPEG-XS is better suited, with comparable high quality and latency in the low single-digit millisecond range. There is, however, a small bandwidth penalty over JPEG-2000, meaning JPEG-XS uses a little more bandwidth for a given picture quality.

UHD and HDR adoption

The pandemic did slow this down a bit, especially in sports events. We see that there is an adoption of 1080P HDR and then upconverting as an alternate to producing in 4K, and that seems to be getting increased visibility.

IP migration

We’re all aware that, at least in long-distance networks, the network infrastructure has been IP for some time – some type of Ethernet layer 2 or layer 3 or MPLS network that is carrying the media content. That is pretty much well established. In the local metro markets, especially for last mile connectivity for media, there is still mostly dark fibre, patched SDI or SDI matrix-based connectivity. We see some incentive due to the pandemic and wanting to minimise truck rolls and have fewer personnel in the central offices, that this area is looking to move to a more remotely controllable operation.

There is a need for switched IP-based solutions that use media-to-IP conversion equipment at the edge of the network and then COTS IP switches optimised for media processing in the centre of the network. Control is similar to an SDI matrix switch with X/Y type (input-to-output switching) control panels for operator ease of use. This gives facilities the capability to move to a more controllable solution where they have good assurance (both equipment and services) and can manage their connectivity without having to have people on site or in their central office.

Software capabilities for remote production

We see our customers sending as few people as possible to venues; we even see this with camera operators located at the studio rather than at the venue running pan/tilt capabilities, something that requires a rather low latency path to the venue that allows them to operate. You can also take systems in the field and through software upgrades that can be done remotely and on-the-fly, set up this capability for different codec configurations (eg, higher quality, lower latency and/or lower bandwidth).

For local area video we are also seeing the switch capabilities managed remotely, so when you need to establish services between locations, you do not need to deploy people. This can literally be done from someone’s home so long as they have a secure connection into the network management system and can then operate their specific links, observe their provisioned links, trouble-shoot or whatever it may be. We see this transition to a software and controllable ecosystem being much more important right now when the customers and users want to put so few people into the field.

With so many choices of codecs, it can be challenging where every event or user is potentially using a different style encapsulation codec. Even when we talk about reliable transport over the internet there are different protocols and different practices. What we see right now is that JPEG-2000 for high end events is mostly the standard and we’re seeing new applications that will go to JPEG-XS for latency considerations as previously mentioned.

For the lower bit-rate content, H.264 has been very strong. It is widely available as a technology and the licensing seems to be more agreeable with most users.

here are a lot of choices now, but latency has to be considered. We’re going into more of a software world where the platforms we’re bringing to market and putting in front of users can have software run different versions of the codecs with their various tradeoffs, whether it be for higher quality, lower latency and/or lower bitrates; the users have the ability to make those trade-offs.