As discussed in the November 2021 issue of Newswatch, interest in media in the cloud workflows has rapidly accelerated, causing great excitement and new options for content creators, while simultaneously posing a series of technical- and business-related challenges for the industry to solve as it proceeds with this ongoing transition. About one year ago, SMPTE launched its Media in the Cloud Advisory Board to begin its process of addressing these challenges, and the two co-chairs of that group—Grass Valley’s Larissa Görner and Deluxe’s Rich Welsh—tell Newswatch that major initiatives are coming in 2022 and beyond.
Welsh points out that the Covid-19 pandemic’s impact on the industry essentially accelerated the overall media push into the cloud because, quite simply, there was no other viable way to get work done for a time. On the other hand, he adds, much of that push was ad hoc and so, now it is time for the development of cloud business models and technical solutions to essentially “grow up,” as he puts it.
“I think the fact of the pandemic and its impact on the media industry, causing a requirement to move from on-premises work patterns to remote working, has certainly brought the whole media in the cloud subject to the fore,” Welsh states. “A lot of companies that previously did not use cloud services effectively had no choice but to move some of their service and operations to the cloud in order to keep going during the pandemic. Therefore, they effectively started their transitions, but in a sort of de facto way. Now that they have done that, they need to decide if they will continue working in a hybrid sense or transition to permanently working fully remote from a facility point of view. So, the impact the pandemic had on the industry’s media in the cloud initiative was to basically serve as an accelerator.
“But a lot of that was done very rapidly, and in many cases, not necessarily optimally. So, that has been disruptive in the sense that workflows have had to change, operations have had to change, systems and processes are all changing, but not in an ideal fashion. So, what I think we’ll see over the course of the coming year is people stepping back a little bit to take a breath and examine how they want to fully transition into the cloud in a way that they want to, rather than as some enforced requirement. Effectively, this is about maturing the approach to how to transition into the cloud, and maybe bring some finesse to it—less duct tape and a bit more engineering, in a sense.”
Görner explains that the Media in the Cloud Advisory Board recognized early on that “there were so many topics within this whole transition, so many challenges and things to adapt to. So, we thought it would make sense to break it all down into smaller groups—what we call focus groups—and we did a survey within the community to identify what those areas should be.”
They identified four topics “that were burning the most,” in Görner’s words, in terms of transitioning media content production procedures into cloud-based procedures. Those topics were live production, supply chain, workflow, and business challenges, and focus groups have now been formed in those areas.
Additionally, a Media in the Cloud Super User Group was being planned at presstime, which the Advisory Board refers to as “an interactive community to learn, share, and discuss challenges in the transition to the cloud.” An HPA Technical Retreat event in late February will bring the Super User Group public, with a breakfast discussion on best practices, according to Görner, before a formal launch at NAB in April.
“The Super User Group is really for everybody,” she says. “It’s really a place where opinions, best practices, and use cases will be shared as we moderate them between people from C-Level, engineering, and operations of production houses, as well as vendors who [eventually], will be providing solutions in this space. For now, the big agenda is to get a critical mass of people participating to make it very successful as we try to solve all these different challenges related to [this transition].”
Welsh points out that much of this work will rely on the use of the 2020 white paper from Movie Labs, a technology joint venture between the Hollywood studios, which as previously discussed in Newswatch, is called its 2030 initiative—essentially a general blueprint for how media companies, particularly in the motion-picture industry, might transition to producing most of their content through cloud-based methodologies. He adds, however, that some of the techniques outlined in that paper need to be more formally tweaked to apply to the larger world of media and broadcast production—a task he says the SMPTE initiative is now diving into.
“That paper probably should have been renamed the Movie Labs 2025 document, since we have accelerated probably about five years of progress in moving toward a cloud-based environment for most media operations in the last year or two, simply because the pandemic forced that to happen,” Welsh says. “Therefore, the horizon has moved a lot closer. But we are working very closely with Movie Labs and others who are active in the Media in the Cloud Advisory Group and our other initiatives to get that done.”
At the core of moving in that direction, Welsh suggests, is the need to evolve the ontology concept cited in the Movie Labs paper to work more broadly across a wider media landscape. This is a high priority for the Advisory Group, and Welsh says a paper on a wider media-centric cloud ontology “is close to publication—something we hope to be talking about fairly soon.”
“The 2030 Movie Labs document is aimed at movie production mainly,” he continues. “And that can bleed over into a lot of episodic production these days, because the streaming platforms are effectively making very long movies in many cases, and their production and post-production processes are pretty much identical to what is done in the film industry. But our audience is much broader because it involves a lot of live television broadcast work and playout services, for example. So, whilst Movie Labs put together their own ontology efforts, we have a parallel ontology program in the Media in the Cloud group, working toward a broader ontology for the sort of whole umbrella that SMPTE covers.”
For instance, Welsh points out that the Movie Labs material includes a suggestion that “all assets are created or ingested straight into the cloud and do not need to be moved.”
“For us, that is still a big challenge right now in terms of bandwidth and production workflows,” he explains. “Also, one thing we are looking to do with this initiative more broadly is help people who are transitioning to the cloud to effectively learn from those who have already done it. But we also want to enable a much more connected environment for the media industry, in general, when it comes to cloud services, so that there is more interoperability and more ability to be dynamic about using client services, and not trying so much to re-create monolithic architectures that we have had in the past—to be more flexible, micro-service oriented, and agile.”
Both co-chairs point out there are a series of issues that will not be simple, quick, or easy to solve in order to move the greater media industry more firmly and permanently into the cloud. As Welsh noted, interoperability between cloud providers who normally service a much wider business world than simply media and entertainment is one of those. Security challenges are another. Tailoring the whole concept to accommodate live production is still another—a concept that Welsh says will essentially require “a big paradigm shift” at the end of the day. Görner elaborates that “cloud-based live production solutions are available today, but media companies need guidance” for revising and evolving their business models for this kind of work. That’s one reason part of the Advisory Board’s work will delve deeply into business strategies and models—an area that, in the past, in other areas, SMPTE initiatives normally did not typically pursue.
Among the crucial business decisions media companies would have to make is whether or not to hire major public cloud providers like Amazon, Google, Microsoft, or others to become foundational pieces of their own cloud initiatives, or whether to attempt to build private cloud systems of their own. Welsh says, for the time being, for most companies, “you would have to make a pretty compelling case to not use the public cloud.”
“That’s simply because of the economies of scale—it’s quite expensive to build anything of that level yourself,” he says. “Media is a classic example of this problem. If your workflows are very peaky, or your demand fluctuates heavily, then really the only way to be able to cope with that is to either build your own setup that lives at your steady state minimum level and then sort of ‘burst’ to the Cloud as needed, which would add more complexity, or simply work in a public cloud. The problem to address with that is, unless you start with a kind of multi-cloud architecture plan, you may well end up in a situation where you really are ‘stuck’ in one particular provider’s cloud. That may not be a bad thing—it depends whether they offer everything you need. But from the cloud provider’s point of view, I think we need to see them offering more native capabilities to media companies, to increase interoperability.
“You have to keep in mind that, from a cloud perspective, your technology refresh cycles will be fairly rapid, because this is all about software, and software doesn’t last that long before you have to update it. In such updates and technology refresh cycles, you can go from a single cloud provider architecture to something that is more adaptable to allow you to run the same services in a lot of different environments of different large public cloud providers.”
Among other things, another one of the big challenges in helping to guide the industry toward a ubiquitous media in the cloud future is the fact that the nature of the technology and its processes do not lend themselves well to traditional standardization processes and timelines.
“Traditional standards do not necessarily fit well with the kind of software development ecosystem that [cloud production] has,” Welsh says. “The [cloud universe] is much oriented toward open-source collaboration and co-sharing. The software world is not static—it changes far too rapidly to fit the simple model for [most standards]. So, whatever emerges, they may not be standards in the way we would imagine a typical SMPTE standard looks like.”
Therefore, Görner feels that while standards in certain areas related to cloud-based media creation may eventually appear, the current need is more for helping devise best practices and meaningful guidance for the industry.
“Today, we need best practices more than anything,” she says. “One discussion we have had is that media cloud processes need to work differently—we need to work on short-term deliverables and let’s say the 10 big challenges of working in the cloud right now. It’s really an iterative and agile process that we want to create within our group, which is a bit different than normal standards’ work. Outcomes will happen much faster.
“But we have to remember that we are not talking only about technology changes. We are talking about full organizational and business changes, and that is why we started our business challenges group. SMPTE has not always been in that space, but with this topic, we think it is essential. We are trying to put this into a market problem metric. What are the challenges we see today and what are the problems that will come down the road, and how to we plan to deal with each of them?”
For all these reasons and more, the co-chairs add that the media industry is transitioning its business practices at very different rates, depending on the organization. Indeed, Welsh emphasizes that one of the big initial challenges is “why move to the cloud at all—what is the business benefit?”
The Advisory Group is working on answers to such questions, but both co-chairs remain confident the industry will, at the end of the day, have no choice but to push ahead. After all, as Welsh puts it, “the advantages will allow us to go much further than we can today. Artificial intelligence, machine learning, automation—all sorts of tools that can help you be more efficient—are there. You will be able to get a deeper understanding of your own content and deploy things like adaptive workflows—things you can’t really do in any other environment.”
In any case, however long it takes, the co-chairs emphasize that SMPTE is off to a solid start in playing a key role in this industry-wide paradigm shift. Görner adds that 2022 will therefore be “a huge year” for cloud-based media creation. “We have learned so much in the past year and see many media organizations ready to take the leap,” she emphasizes.