Broadcasters are already playing out many linear channels from integrated channel-in-a-box solutions. With the recent introduction of IP-based playout solutions, comprised of software running on commodity IT hardware, they now face the daunting challenge of determining the best path forward to meet their operational and business needs.
“The challenge is to create software that sits on generic IT infrastructure and still be as powerful as today’s hardware-based product in order to make the switch economically attractive,” explains Steve Plunkett, CTO, broadcast and media services, Ericsson – a global service provider which runs its MENA playout operation out of Twofour54. “We know that making software that can run on generic infrastructure is not easy so it’s not a surprise to us that the industry is taking time to commit.”
In many cases, vendors are taking product in fixed hardware configurations and deploying that in virtual machines only to find that it needs some serious software remedy to allow the product to operate in the much lesser known IT environment.”
Plunkett goes so far as to suggest that “the marketing messages appear to have got out of step with the engineering reality.” It’s a feeling supported by playout provider Deluxe Media Cloud whose MD, Maurizio Cimelli, says, “Sadly, there is a lot of smoke and mirrors spread by vendors who claim to have product that works.”
Igor Krol, COO at software playout vendor Veset, believes that one of the misconceptions in the broadcast industry has been in defining the cloud as nothing more than a remotely accessed virtualised server in a data centre.
“Virtualisation and the cloud are separate technologies, albeit with similar goals – to provide a way to run multiple software processes on the same hardware,” he insists. “Virtualisation does not deliver elasticity and pooling comparable to cloud technology, and therefore demands a substantial commitment of IT and engineering resources to manage virtualised applications.”
He suggests that vendors who have to adapt proprietary and hardware-based systems for virtualisation face a challenging task, one often still requires specialised servers. “This approach does not allow users to extract the maximum benefits from ever-falling cloud resource costs,” he says.
A Disaster Approach
Service providers stand to benefit most from the move to cloud playout because of the scale of their operations. “It’s an evolution, not a revolution,” says Tom Gittins, Sales Director, Pebble Beach Systems. “Broadcasters are highly risk averse and will probably start to use this technology in lower risk deployments such as Disaster Recovery (DR).”
This means that valuable content is typically stored at a second site with media played back on air through an IP output derived from a virtual environment within hours in the event of failure of the main channels.
“We recognise the need for a period of coexistence for best-of-breed deployments of classic technology alongside the adoption of IP and virtual machines,” says Gittins. “DR deployments will open up a range of options for traditional channels which may currently run without a DR plan.”
While there are advantages in the move to cloud of lower operating costs, simpler monitoring and management, it is the ability to turn on and off channels as needed which is seen as the game-changer. The ability to rapidly change channel characteristics, to launch channels quickly, and to launch experimental channels will allow broadcasters and content owners to compete with the flexibility of online only networks like Netflix.
“Knowing you can provision a channel days, even hours, before it’s needed, change its available features through software license keying, and utilise resources from an available pool rather than dedicated devices are highly attractive features of the cloud-based approach to playout,” says Andy Warman, director of product line management, servers and storage, Harmonic.
Conceived this way, ‘publishing’ may be a more appropriate new term for the process than playout. “We are helping clients shift their thinking from channels to publishing streams,” says Imagine Communications’ VP of playout, Tim Mendoza. “Streams offer totally different ways to connect to customers. They can be personalised so you can sell more advertising with digital ad insertion and therefore make more revenue from content.”
The opportunity for service providers is to become the expert in these types of solutions. Their business becomes, says Warman, “less about the management of channels and more about management and orchestration of the IT resources” that can host the shifting needs of broadcasters and content providers.
A service provider’s business is no longer about linear, but all the other associated and ancillary services such as VOD, live to VOD, OTT and enabling the personalisation of content on any and every platform.
“Virtualised platforms, possibly in the cloud, are enablers to all of this brave new world and it’s where the future is heading and needs to be embraced by those in the service provision game,” says Snell Advanced Media’s director of playout, Karl Mehring.
Nonetheless, there are costs associated with a move to cloud. “It’s important to understand that if the service provider wants to build their own private cloud from scratch they will have to make substantial upfront investments,” says Krol, who puts this in excess of US$10m. “Then they need to maintain an expensive team of engineers to look after it and have a large portfolio of clients, which can utilise those capacities. Just putting up a rack of few servers in an air-conditioned room does not make that datacenter cloud-enabled.”
There is widespread acceptance that virtual machines are the way forward, but questions remain over on air reliability, and there are some factors which are definitely encouraging a wait and see approach, not least 4K.
UHD A Barrier
UHD playout of any kind is in the early stages of rollout. Even with conventional solutions, it is not an easy task to put UHD channels on air. Technical aspects as asset bitrate, file size, or IP transport of the UHD feed do not pose a challenge, according to Krol, but RAM speed and the amount of CPU cores required to process and playout UHD content do.
“The industry’s drive to HDR, 10-bit, higher frame rates and other approaches to create a more immersive experience for the viewer are also adding to the complexity of UHD,” says Warman.
According to Mehring, the financial costs of storage and time taken to move the very large UHD file sizes around “are going to make VOD the most attractive initial delivery mechanism for single assets and binge viewing.”
However, all are confident that as bandwidth costs come down and compression techniques improve it will be just a matter of time before UHD becomes commonplace in linear playout, whether from the cloud or anchored to the ground.
This applies to branding using complex graphics which any high value channel depends on. According to Warman, “CPU, rather than GPU, graphics are up to the task in all but the most demanding applications today. It most cases, it is not required to have GPU capabilities.”
Krol agrees; “The overwhelming majority of existing linear channels can be serviced by cloud playout using standard CPU-based processors available from public cloud providers.”
A Phased Migration
Service providers, taking the cue from their client broadcasters, will not make the move to cloud in one big bang.
A phased approach involving less operationally demanding, less complex and less reactive channels or in which premium channels are mirrored for disaster recovery will be moved first. Once a particular type of channel is templated and assuming pre-integration and testing, then the time it will takes to initialise a new channel need only be a matter of minutes.
“For too long the industry has been overly cautious of IP and IT,” stresses Plunkett. “At Ericsson we tread a mid-ground where we desire to take advantage of it but we want to do so with all the Key Performance Indicators and Service Level Agreements that client’s expect.”