SVG Europe Sit-Down: ATEME VP marketing Remi Beaudouin discusses viewer content expectations and delivery developments
Increasingly, industry discussion is dominated by the move to IP connectivity, whether it be for contribution or distribution. The sheer volume of data to be transferred means that compression efficiency is a critical issue, while the sheer number of content paths means that automatic provisioning, monitoring and quality control of links is vital.
Current forecasts are that video consumption on mobile devices will grow by more than a third in 2016 alone (34.8% according to Zenith Optimedia). One of the major drivers for that growth will be the two big sporting events this year: the Euro football championships in France and, of course, the Olympics in Rio.
Delivery evolution and viewer expectations were therefore high on the list of priorities when SVG Europe spoke to Remi Beaudouin, VP marketing at video compression specialist ATEME…
What is driving technological change in this area?
The underlying message is that consumers want more choice, on any device, all the time. That is leading to real challenges both for creators of content and for those who have to deliver it.
It is not just volume – it is trends which involve bigger data files, like 4K (and probably 8K) video and virtual reality. Finally, consumers are focused on quality. They no longer regard video to mobiles as a miracle; they invest in the best screens even on their phones, and expect the best pictures on those phones.
What is happening in distribution?
As you know, the rise in mobile content consumption is staggering. The 34.8% growth that is forecast for 2016 follows on from 43.9% growth in 2015.
Consumers want to use their phones and their tablets to catch up on favourite programmes, but most of all they want to use them for additional content. In sport they want to be in the locker room before the game, sitting alongside the coach and seeing the decision-making processes during the match.
As 360˚ video becomes a reality, so viewers want to really feel a part of the event, to be right there. All these technologies are now ready and mature – but all require more data to be prepared, encoded and delivered.
What is the core challenge for delivery?
We already know that the Euro football championship with see at least eight games covered in 4K/ Ultra HD (UHD). That coverage will be taken by broadcasters in France, the UK, Turkey and further afield in Korea and Japan, and probably many more.
Where 360˚ immersive video is available, that will probably be the equivalent of 4K resolution, too, so you need to plan for at least 15Mb/s to each user. And, as I say, this has to be a good and consistent quality of experience; consumers are very intolerant of poor video.
The solution for delivery must lie in virtualised, software-based solutions. It needs an all-IP solution, running on COTS hardware in a software headend, and perhaps even in the cloud.
This is the only solution that will deliver flexibility. Modular software gives you the agility to deliver what is needed: any input, any codec, any output, any resolution. CPU-based solutions are also inherently elastic, allowing dynamic allocation of resources. A software headend can make intelligent decisions about absorbing peak loads when needed, or enhancing video quality when CPUs are available.
Central to this is the migration to HEVC. Software encoders are now starting to deliver the promised benefits of this radically different codec, reducing bandwidth over MPEG-4 by 30 to 50%. Quality is increased; more signals can be handled; and UHD becomes cost-effective.
And looking at business issues rather than technical ones, a move to a software headend shifts the finance from capex towards opex. The capital costs are reduced because broadcasters and service providers can buy their own COTS hardware; flexible licensing links opex to outputs much more directly, ensuring transparency in understanding the costs of each service.
Is it all about online consumption?
I think consumers will still want broadcast television, not least for the big sporting events which are social occasions with people grouping to watch together. So we need to find better ways of delivering more content and alternative content over conventional platforms.
One of the most important advances is DVB-S2x, the new extension to the satellite DVB standard. It increases modulation efficiency by 15%, which in turn generates significant savings in cost per bandwidth. Our Kyrion encoders and decoders are already DVB-S2x-ready, and allow satellite operators to add services within their current contracted capacity, effectively reducing the opex for each service.
How do advances in IP delivery affect production?
The big trend this year is something we saw starting two years ago with the Sochi winter games: remote production, or the stay-at-home production model. Broadcasters bring back multiple, uncompressed or lightly compressed feeds over fibre, then edit the stories in their own premises. Staff are more productive because they are in their familiar, comfortable environments, as well as saving travel and accommodation costs.
Advanced remote production projects might have as many as 45 feeds over a 1Gb fibre, with full remote control of the cameras from HD. Variants include editing on proxies back at base with all content stored on site and only the selected clips transferred over the fibre.
For this to become a real sports solution, particularly for full remote production, then you need consistent, ultra-low latency. ATEME’s solutions provide links with a maximum latency of around 150ms, which is consistent across all links, making realtime live remote production a practicality.
Is it practical to use the public internet for broadcast-grade circuits?
This is another area of interest, particularly where broadcasters and production companies are looking to control costs. The challenge here is that you have to find a way to maintain the quality of service when you have little or no control over the actual circuit.
One way of achieving this is with a new standard called Automatic Repeat request (ARQ). This is an error-control method that uses acknowledgements between the encoder and decoder. Should a packet get lost on the open internet (or a network without QoS) then the encoder will quickly re-stream the packet, repeating it until it receives the acknowledgement or it exceeds a pre-defined number of attempts. There is no point in adding latency unnecessarily to recover from the rare case of a lost packet!
Have you a recent major sport project you can describe?
We are not able to talk about the user or the project in detail, but [we can briefly outline what we have] delivered for a major player in sports broadcasting. At the heart of it is a dense blade server providing flexible realtime encoding and storage management. This gives a very fast throughput for multiple feeds from remote production sites, with access for fast turnaround editing on growing files, as well as broadcast and OTT outputs and archiving.
Because this is a software-defined system, it is totally interoperable with other software-defined solutions such as playout and the non-linear editing applications. Implementing it in a virtualised environment means that load-sharing, redundancy and resource allocation is completely flexible – a practical solution for the fast-moving, fast-changing world of sport.