The quest to build and integrate cloud-based workflows for real-time live production has begun, write Hans Hoffmann and Ievgen Kostiukevych. The EBU will ensure broadcasters get an edge as the ecosystems evolve.

COVID-19 caught the world off-guard. All industries urgently needed to come up with solutions for business continuity. For broadcasters, this meant finding remote and distributed production solutions. And for most of them, the cloud, in one form or another, was the answer.

The real-life proofs-of-concept that resulted have demonstrated clearly that using the ‘public cloud’ for remote and distributed media production, and even live, is a genuine value proposition for the future. And while some broadcasters have opted for ‘cloudifying’ their existing workflows, others have successfully started using the growing portfolio of native media-tailored services from cloud providers. These first solutions have not been perfect – far from it – but they have provided quick and effective solutions under highly unusual and difficult circumstances.

STUMBLING BLOCKS

It quickly became apparent that most of the in-studio standards and protocols currently used are not designed for cloud-based systems. The architects had to evaluate the connectivity type (public internet, direct, or WAN connection) and, depending on the options, consider link properties like latency, reliability, security, and bandwidth requirements.

Another stumbling block was that the public cloud is asynchronous, both historically and by nature. Imaginative hacks were required to enable synchronous and time-sensitive workflows. Public cloud providers are only now starting to investigate offering multicast and PTP for their customers. While the offer of native media services is continuously expanding, and synchronous workflows may be enabled in the future, we must nevertheless ask what portion of our production remains truly time-sensitive? Where can compromises be made in favour of greater flexibility?

Up to now, broadcaster adoption of IP-based technology for live production has focused on the ‘ST 2110 plus NMOS’ ecosystem. Though the complete protocol stack isn’t quite there yet, early implementations have provided confidence that it will be the go-to solution to offer reliability and flexibility for open high-performance uncompressed infrastructures. ST 2110 plus NMOS provides a solid basis for on-site time-sensitive infrastructures, like local MCRs and local IBCs. However, for high-quality live production, uncompressed UHD/HD bit rates can quickly make the use of cloud-based systems impossible. There is a clear need for new mezzanine bit-rate compression solutions that will ensure manageable bandwidth and costs, even for high-value high-quality cloud-based or hybrid productions.

NEW STANDARDS?

There is a lot going on. Some of the big industry players are investigating ways of retrofitting existing solutions into a ‘cloudy’ future. In parallel, SDOs have launched initiatives to enable time-sensitive cloud applications in a relatively seamless way. Examples include AES67-over-WAN from the Audio Engineering Society or the ST2110-WAN and GCCG (ground-cloud-cloud-ground) projects from the Video Services Forum.

SRT (open source) and RIST (standardized by VSF) both solve at least one of the essential puzzle pieces in contribution applications, targeting video transport at low latency over unmanaged networks. Although both are asynchronous, they are supported by major cloud providers either as a native offering or with minimal integration required.

The change that is now under way will be even more dramatic than the transition to IP initially seemed five years ago. New paradigms need to be adopted – and very quickly! The cloud giants are investing much more into R&D than a union of broadcasters can afford. They will soon take the media industry by storm with new offerings that are not built on a legacy broadcasting mindset. We will have to live with that, adopt them, and use them to our benefit or, in the longer term, be bold with a European counterpart cloud like GAIA-X.

In any case, we need to establish constructive dialogue as a collective with cloud providers. By giving them early insights on what kind of solutions work and which requirements are still to be addressed, there is an opportunity to shape these upcoming offerings.

Learn about and help shape future cloud workflows

Employees of EBU Members are invited to join the working group on 'Hybrid and Cloud-based Production'.

Latest news