Live sports streaming can deliver huge benefits – both to service providers and their customers – but it comes with big challenges. Synamedia’s Simon Brydon and Marc Baillavoine discuss the merits of streaming and how to overcome the hurdles.
Synamedia demonstrated a range of live sports solutions at this year’s NAB Show
Why does it make sense to stream live sports rather than rely on tried and trusted broadcast technology? After all, broadcast technology is highly suited to delivering mass-audience live events to millions of viewers simultaneously. Its one-to-many capabilities are unmatched. However, there are several reasons why streaming technology is now seen as representing the future even for events with millions of concurrent viewers.
One is that consumers are increasingly watching sports on devices that can only be reached by streaming – their phones or laptops. And sports fans may also view games on the big screen by casting a stream from a phone or other device.
A second reason is that streaming has global reach, enabling sports bodies to aggregate audiences, particularly for niche sports that may not have a massive audience in one territory. Streaming also enables sports streamers to pay as they grow to some extent, without expensive upfront commitments to satellite capacity, for example.
Compared with broadcast, streaming can also deliver a superior user experience. It more easily enables instant access to multiple camera angles to provide a different perspective on the action, for example, or it can be used to enable watch parties to enhance fan engagement with services.
An ancillary factor behind the growing popularity of live streaming is that sports bodies want to enhance the experience of fans standing in a stadium watching, say, a live football match, by enabling them instantly to call up replays of key moments such as goals and statistics about the scorer.
More broadly, the overall direction of travel in video distribution is towards IP and towards adaptive bit-rate delivery, enabling all screens to be reached via one technology.
Viewers, of course, accustomed as they now are to viewing content on different devices, do not really distinguish between delivery mechanisms – except when things go wrong.
The challenges – bandwidth, latency, piracy
And for mass-audience events, things do sometimes go wrong. The unpredictability of demand, sudden surges of traffic at key moments in a match, millions of viewers trying to register simultaneously – these are factors that can cause live sports streaming services to go into meltdown.
Managing bandwidth cost-effectively, preventing buffering and freezing while keeping latency down to a respectable level – the overall quality of experience – and piracy. These are the big challenges most often cited.
For big sports – Premier League football, the World Cup, the Olympics – the challenges are amplified.
“Sports streaming needs to be at scale – that’s what it’s all about,” says Marc Baillavoine, Senior Director of Product Management, Video Network at Synamedia. “That also means instantaneous scale. There is nothing before the event and then suddenly everyone is on the pitch. That kind of streaming at scale is very difficult.”
Adequately planning and provisioning for demand is extremely challenging. Service providers need to over-provision to avoid the onscreen action juddering to a halt in a way designed to enrage fans and potentially destroy streaming services’ reputations. But they can’t spend infinite amounts of money to ensure that nothing goes wrong.
Typically, sports streamers are likely to use a mix of pre-provisioned CDN capacity, with payment based on time, topped up with capacity to meet unexpected peaks in demand that is paid for on a per-volume basis. Getting the balance right between this is important to keep costs manageable, but it is to some extent a guessing game.
“One problem is the idea that for the CDN to deliver your sports content at quality, you need to size it for the worst case. You may only need that capacity for 5% of the time. So you can use the cloud to offload when you don’t need the capacity,” says Simon Brydon, Senior Director, Security – Sports,Media & Entertainment,Synamedia.
Quality vs latency
Service providers must also consider the balance between quality – the expected resolution and how far this can be degraded to ensure a consistent experience – and latency – how much of a buffer is needed to ensure that consistent experience without forcing viewers to become aware from social media that a goal has been scored before they see it themselves.
In any contest between reducing latency and ensuring the image does not freeze, no freeze is bound to win. Streamers record far more complaints about frozen images and poor quality than they do about latency. But there are nevertheless limits to tolerance for latency that service providers need to consider. Latency also becomes important when service providers want to introduce applications such as multi-angle viewing and watch parties.
“When you are streaming at scale, latency is a different order of problem,” says Baillavoine.
While there are streaming protocols available that can ameliorate the latency problem, deploying these can also lead to other issues.
“Most of the solutions that deliver low-latency streaming are based on WebRTC, which is not very secure and doesn’t scale very well,” says Baillavoine, suggesting that streamed events that reach over 100,000 people would be “already challenging” for the protocol.
The ultra-low latency enabled by WebRTC may be necessary for niche applications such as real-time sports betting, but it clearly has shortcomings in this environment.
Instead, suggests Baillavoine, those streaming sports events at scale would be advised to look to low-latency DASH or HLS (DASH-LL or HLS-LL) – extensions to commonly used protocols that may be more suitable for streaming media at scale.
“The problem now is that you must control the whole ecosystem, including the CDN. It’s very difficult for the media owner that doesn’t own the CDN to make sure the content will be delivered with low latency,” says Baillavoine.
Streamers without this luxury, who are forced to rely, at least in part, on public CDN infrastructure, he suggests, “need more tailored solutions”.
However, he adds, how far streamers want to go with this “depends on what you are trying to achieve”.
If it is acceptable to deliver a sports streaming service with “broadcast grade latency” – meaning eight to 12 seconds – “then a public CDN will work”.
On the other hand, streamers that are seeking to differentiate the user experience by reducing latency further, for example to achieve the gold standard latency of two-to-five seconds, will require much more precise control of the infrastructure over which the video is delivered.
“Deploying low-latency encoding is a good first step but if you don’t control the player or the CDN, that’s a problem,” says Baillavoine.
In practice, streaming operators may wish to compromise in order to maintain control over another key aspect of making sports streaming work – costs.
Solutions are available to enable advanced features by synchronizing latency at an acceptable level. High-efficiency Stream Protocol (HESP) can enable streams to be synchronized across devices with as little as two seconds latency. And DASH-LL and HLS-LL can deliver broadcast-quality latency if streamers have control of the CDN, with AI-enabled features added to predict scalability when traffic levels vary.
For up-and-coming 5G-enabled services such as enhanced in-stadium viewing, meanwhile, tailored solutions are available. At this year’s NAB show, for example, Synamedia highlighted its ability to deliver viewing in real-time via different camera angles, along with instant player statistics, on 5G phones in partnership with NativeWaves. The application featured Synamedia Vivid compression,Vivid OTTwith ultra-low latency and fast channel zapping, and secure delivery usingFluid EdgeCDN.
The piracy threat
While there are technological fixes to some of the delivery hurdles, streaming comes with other challenges that can have a more direct negative impact on revenue – notably that of piracy.
“There has been a massive rush to grow subscription services and platforms have been built without very strong inherent security,” says Brydon.
“Those businesses are under enormous pressure to grow and quite often security is seen as a less important element. So organised criminal pirates have perfected the art of stealing content from legitimate broadcasters, hacking CDNs and breaching what are in many cases very vulnerable systems.”
With streaming technology now available to a wide variety of people and organisations, it has become much easier for criminal groups to launch illicit services that aggregate content in a way that is closed to legitimate players and deliver them with a professional user experience.
Brydon says that there is “clear evidence” that those consumers who access illicit services are also the biggest consumers of legitimate offerings, meaning that they are probably for the most part aware when they are consuming illicit offerings.
“For the industry to counter that it needs to look at the economic model. On the technology front they must protect their own networks and content in a way that forces the pirates to distribute substandard video content. They need to do more with technology, which in the case of live premium events means forensic watermarking,” says Brydon. “That enables them to triangulate interconnected networks and identify sources of leakage.”
While the issues around latency and quality of the streaming experience can potentially make or break a sports streaming service, discussion of piracy – which presents a massive threat to sports services – also raises one of the other key challenges facing providers: how to monetise the offering in a way that secures a viable return on investment. The migration from legacy broadcast to streaming has gone hand in hand with the breakdown of the model of an all-in pay TV bundle and a move to a more à la carte way of consuming services. This brings its own challenge – how can a single-service streamer faced with significant content costs be sure of delivering a viable return on investment?
For services that acquire premium rights, there is no substitute for the subscription model – however challenging that may be in the new world. But interest in tapping into the global reach of streaming to monetise through addressable advertising is also growing – particularly for mid-level sports that don’t quite match the appeal of, say, Premier League football.
“Many niche sports with a diehard fanbase will attract paying customers. But other mid-level sports require a degree of mass distribution. There is an issue for OTT sports in that a big part of the revenue for the sport may come from advertising, but not addressable advertising – it comes from sponsorship,” says Brydon.
Sports streamers must calculate whether it is more profitable to opt for subscription, and sacrifice advertising revenue, or go free-to-air to secure an audience that might attract sponsors, and supplement that revenue stream with addressable advertising, he says.
It’s a difficult balance to get right. But for the technical and commercial challenges – bandwidth management, scalability, latency, security, subscriber management and monetisation – streaming delivers massive benefits. These include the ability for services to pay as they grow, the benefit to consumers of accessing advanced features and the ability to seamlessly follow the action on multiple devices, with multiple viewpoints. For all of these reasons, the rise of live sports streaming seems unstoppable.
This is sponsored content.