Mark Turner, Project Director of Production Technology at MovieLabs, chats with Michael at NAB about the 2023 Vision and their goal of bringing a more efficient workflow process to the entire industry.
Michael: Hi, Michael Kammes here from Shift Media, booth N1875, and on the show floor today, we're talking to folks who make a difference in our industry. And today, we're talking with Mark Turner of MovieLabs. Mark, thanks for joining us today.
Mark: Thank you. I'm busy with making a difference. It's good.
Michael: Excellent. So MovieLabs. There has been a lot of buzz about it. I'd like for folks, the uninitiated folks, to kind of understand what MovieLabs is. Can you explain to our folks out there what MovieLabs is?
Mark: Sure. MovieLabs is a joint venture, technology joint venture, of the major Hollywood Motion Picture Studios. And we've actually been around for 15 years now. Started out in distribution technology. We've been moving up the stack and now are doing a lot of work in production. That's where people are hearing about it now, the 2030 Vision. There's a MovieLabs paper that came out in 2019, which is very focused on where do we need to get to to make a more efficient workflow process. Because let's face it, our industry is not scaling well, and we are gonna be asked to do more and more scaling, and our old a hundred-year-old workflows are gonna fall apart. So we need to find a better way of doing stuff. So the studios, through MovieLabs, have put out their paper. It is the common studio viewpoint.
Michael: They got a whiteboard and said, this is where the industry is going. This is where we need to go.
Mark: This is where we should all get to, not just for the studio's benefit, but for everyone. All boats will rise, right? We will not be able to scale this industry if we don't sort of get to this place. And it's very cloud-based focused. But there's a lot of work to be done, which is why we blew it out ten years. There's workflow implications. It's training implications. Like, this is a big change, and it's the change we never really did when we went from analog to digital. I mean, we went from analog to digital, we changed the cameras, changed from moving things in film cans to hard drives, but the workflows really didn't change. And we're doing a pivot into the cloud. That opens up a whole bunch of new technologies. Like, this is our chance now to change the workflows and really get a more modern media workflow system running. And if we don't do it now, we're gonna miss the boat.
Michael: I love the organization of the paper because it was actually broken out into many principles, ten different principles. And we'd be here all day if we talked about all the nuances of the ten principles, so maybe you can make that a little bit more digestible.
Mark: Yeah. So the ten principles, they break down into three areas. That's an easy thing, right? The first five are all about cloud foundations. So the first couple really talk about the idea that all the media created coming outta the cameras should go straight to the cloud and then stay there, which is critical, right? Because what we do now is we create media, and then we send it to someone, and they do some work, and then they send it to someone else, and they do some work. They send it, send it, send it, and they duplicate it and copy it. And we ended up with this proliferation of content, and we need to stop that. So the idea is we put everything in the cloud, there's a single source of truth, and then people come in and remotely work on that.
So, you know, the second principles talk about applications coming to the media instead of the media moving to the applications. Media is big. Applications are small. We should stop moving media around the place and move applications to the media. And there's some stuff about archiving in there, but the first five are all about this sort of cloud foundation. And then on top of that, we've got security and identity, which is this idea of, okay, now we've got everything in the cloud that's inherently connected to the internet, bet I could fix security while we're at it. Um, so there's this specific work we've been doing in our security, and that's foundational as we did that work first. And then on top of that, a cloud foundation, a good security platform, now we can start talking about software-defined workflows and getting really interesting, clever abilities to automate things, move things through the pipeline faster.
It's metadata in there. There are ontologies in there. There's the ability to sort of create more interesting pipelines on how applications are talking to each other. So we have to do all of them together. It's not like we're waiting, you know, until 2029, and we're gonna drop this solution on the world. We have to do the work now. That's why we're here. There are companies right now that are demonstrating parcels of 2030 Vision, like, right here today, which is great. You know, SaaS platforms like this - great, we're all for that. We're all for open AI, APIs, and, you know, the ability to interconnect things. So that's what you know we are here for, is what bits can we do now so that we can check them off, and we're looking for the gaps. What still needs closing, and where does technology or training or some new processes need to be fixed? And that's what we are here to do.
Michael: That brings up a really good point because the paper was originally released in 2019, and since then, it's felt like a decade because of various things that have happened in the world. But I'm sure over the past almost four years, there've been some things that have needed to be updated. Right? So there's been some updates from the 2030 Vision released in 2019.
Mark: There have been new releases. We haven't changed the principles, right? The principles are fine. Someone asked me last week, have we changed the principles? I was like, I don't feel a reason to change the ...
Michael: No asterisks at the end.
Mark: No. Because they were aspirational anyway, right? They were big, and they weren't specific. And MovieLabs does not dictate technology to anyone. You know, the studios can't and should not be doing that. What we do is set direction and say, if we all got here, could we make a better world? So, no, we haven't felt the need to change anything.
Michael: But there've been updated papers.
Mark: Yeah. We're releasing new content and doing work as well. We're actually now, you know, writing code and actually deploying tests and putting different companies together and making things happen. So yes, we've put out a paper on our software-defined workflows, which defines that concept and then goes deep into what's required. We've done a lot of work on security. So the last part of what we call the Common Security Architecture for Production, which is a five-part architecture...
Michael: That lower third for that is gonna be unwieldy for you!
Mark: See, we call it CSAP. That's easier. Add that. So CSAP part five just came out. There's one more that'll come, but that takes this sort of cloud well-understood security concepts that work, IT people use 'em every day, right? Cloud is used for government. It's used for military. Like, the cloud is a very inherently secure place. We can make that work for media workflows. We just need to sort of approach things slightly differently in our heads. CSAP is about, how do you take sort of well-established, current technology that works in the cloud and apply it to production? So it's a whole architecture. People pick it up, use the technologies today and actually build a new security system.
Michael: That's actually pretty interesting. You said people just pick it up, and I think what's interesting is because you're putting out these principles and putting these kind of guidelines, it's where folks should end up going. The 1,700 manufacturers that are here at NAB, how do they approach MovieLabs or how do they say, "You know what, we like where this is going. How can we get on board and contribute to that?" What is the usual process there?
Mark: So, you know, MovieLabs are not a standard-setting body. We're not out certifying things. The paper is public. It's designed to be out there. People can read it. People can download it. We have got companies that are writing blogs about what it means to them. Google this week just published something about how do you take the CSEP principles and apply it on Google Cloud today. Amazon just finished a three-part blog series about exactly the same thing. How do you take that security thing and map it today to stuff you can buy off your Amazon marketplace? So any company can get involved.
Michael: But there's gotta be common directions and not just the principles.
Mark: And that's all that MovieLabs is doing. Right? We are just funneling everybody in the right direction. And the Vision is a vision. It's a roadmap. It's like, here are the things we think we have to get done. But we're not building products. We won't build products. We're not for that. We want everybody else to go and build the products. So we're just gonna make sure that it, you know, if you're gonna do that, could you do it in this way so it will work with other things upstream, you know, or downstream. Or we can pass data backward and forwards and stuff because we're looking at this bigger picture than any one particular part. But yeah, the more implementers, the more vendors, they can ping me, look it up online, follow the blog series and follow us on LinkedIn. That's my biggest thing to people.
Cause there's constant new thinking coming out of MovieLabs. And we're also running this showcase program, which came out of IBC last year, where we're actually taking case studies. We're working with the companies who implemented a solution that demonstrates the principles today. And then, we posted them on the MovieLabs website. So, you know, you can go and look right now, and you can look up archiving use cases from Disney, which they did with Avid. You can look at interesting workflow things that Skywalker have done, and you can say, all right, I'm interested in how do I build a new security workflow ontology system, and we found someone who's already done it. We're gonna work with them, write a case study, and publish it online. Like, all boats will rise if we share our knowledge.
Michael: I love the fact you specified that there's no certification. To be very transparent about that. There's another company that also doesn't do certification but is very important to our industry, and that's the MPAA and their TPN+. That program just came out, and I'm curious if there were any discussions or back and forth on how the two bodies may work together. Because the TPN+ obviously is making sure that things that are being looked at, someone's facility, kind of adhered to some of the best practices of the industry, which MovieLabs is obviously influencing.
Mark: Yep. I mean, so we know the TPN goes very closely...
Michael: But there's no TPN certification. They're very clear on that. There's an audit.
Mark: There's an audit, and we won't do that. So we would look to organizations like TPN to go and audit. So whether someone's done a good version of the MovieLabs architecture, ours is an architecture. We would hope everybody implements it in a good way. But we're not going to get measuring. That's not the role of MovieLabs. That's what TPN is for. And that's great. That's, you know, we'll get to a point probably in the next year or two where they can look at what we've built, or what we're proposing, and they can look at that. How do you map that into either TPN+ or a different version of it? That's a TPN question.
Michael: I'm sure we're all tired of talking about the pandemic, but some technologies obviously were accelerated during that process. So did you see gaps or similar other technologies that just, wow, we didn't see that coming during the pandemic that has since caused some of these different papers to be updated?
Mark: No. Someone suggested that we caused the pandemic to try and prove that a cloud-based workflow would be a good idea. I would like to put that to the bed. That is not true.
Michael: If you had that kind of power, I'd like to talk to you after this interview.
Mark: So the pandemic proved a few things. One, it proved that people can actually work remotely, which for a lot of creative jobs, a lot of people were like, "We couldn't possibly do this remotely. How can that ever work?" And then, all of a sudden, in two weeks, they were doing it. Right? So for creatives, it kind of moved a lot of the mindset to, "Yeah, we can do this job remotely," which is great. Because a lot of the Vision has this idea that you can work from anywhere cause everything's in the cloud. So we were kind of there, but we're also not declaring success because we had two years of people working from home. Working from home and the Vision were not the same thing, right? So it was a lot of rush jobs to get people to go and move media to go and work from home, and they were tending to do it in an isolated place and then sent the media back to central office. That was not the same thing that we were talking about. So it moved us a little bit further forward in mindset, at least. But it didn't necessarily fix all the technology problems we needed to get fixed to make a permanent change to an entirely cloud-enabled workflow where everything goes in once, and it stays there and doesn't keep moving in and out and backward and forwards. That's inefficient. We don't wanna do that again. That was, see my earlier point of analog in digital. So progress. It was not a good thing, it was a disaster, but we've pivoted to make it at least learn some lessons out of it.
Michael: In terms of progress, I wholeheartedly buy into the Vision, a lot of folks here do, but there's gotta be, I don't wanna say naysayers, but folks who are digging their heels in. Aside from the obvious, "Well that that'll take too much time" or "That will cost too much money," what are some of the objections that you're hearing from the small group of folks? And how do you typically respond to those?
Mark: I think what we hear from people is it's gonna be hard. Which that's why we gave ourselves ten years. We are a very rare industry where we'll spend a hundred million dollars on a single product, right? And have two and a half thousand people working on that product for 18 months. A lot of whom are freelancers. And then, you know, they walk out the door. You know, if you went to Ford and said, "Hey, why don't you make a new car cost a hundred million dollars, and we'll just make one of them, and then the whole team manufactured it would just break it apart again." And they [say], that's crazy.
Michael: I'd like to be in that pitch meeting.
Mark: But we do that, and that's the complexity, right? We have a lot of people that are not employees that we need to put together. We need a lot of tools. A lot of tools have custom plugins. You know, we have to make that whole ecosystem work in a way that you can swap in and out components, right? And this idea of a software-defined workflow is that I'm not locked into this pipeline. I have to do it this way. You know, we started production this way. We can't possibly make a change. Like, if midway through an 18- or 20-month, 24-month production cycle, someone comes up with a great new technology, you should be able to drop in it. Like, if you say that to someone right now, a producer will go, don't touch anything; it's working.
Software can fix that, but there are a lot of people to be involved. So largely, what we have is skepticism that we can pull this off, which, as we get more momentum, I'm less worried about. And then the other one is change management, which is, okay, you build the best technology in the world, but if no one uses it, or no one knows how to use it, or it's super complicated to use, then we failed. So a lot of my work is actually about, okay, can we bring the people along as we're doing the new technology? So, you know, we are not gonna get to the end of 2029 and say to people, "Hey, there's a whole new solution out there. Well done. Here. Go." And they'll go, "I don't wanna use that. That doesn't do what I wanted it to do." No, don't work like that. So we've gotta get people on technology to move in parallel.
Michael: If we move a little bit more topically, obviously, the big buzzword is machine learning, AI. How do you feel that may influence any of the ten principles or at least some of the updates since then?
Mark: I'll tell you the high-level viewpoint we got of it, and it's mentioned in the original paper, which is that there's gonna be a whole bunch of AI tools. We predicted then in 2019. Some of them are gonna be very useful. What will be the most important thing that we think we can help with is a common data format for them to learn from. And to understand, you know, if you've got a particular data model, who owns that model? Who created that model? A lot of innovation will appear if we can all standardize the data that flows underneath it, right? Then you can do creative things with workflow, and you can plug in different applications. So we built this ontology, the ontology for media creation. Again, it's in multiple parts because it's designed to be extensible, and that's all up on the website.
You can go look at it, but it defines basic terms that we need in our industry that other people don't. You know, we have characters and actors that are related to each other, and they're mentioned in a script. And you end up with these very large data models when you start going, "Well, wait a minute, I've got a shot, and we've got a shot. There was a take, and the take had these actors on stage in it, and they were portraying these characters. Um, but one of them had a stunt double in as well, and that was also portraying the same character." So we get very complicated data models that come, all of which are custom for every production right now. And we think a lot of software and a lot of tools will be a lot more useful if we can actually all standardize the data so that everything is actually being able to innovate on top of that.
So that's a big part. And AI, one benefit from that is that you can actually then have it understand context, which a lot of the excitement right now is about natural language processing in that you can tell it what you want, and it'll go and create it. That still struggles if you don't give it the right context to understand what the language is. You know, if I say "shot" to something at ChatGPT, it might think guns. We weren't talking about that. It doesn't understand the context. So we need to give these tools context we had to make them useful for us. So there's a lot of work to be done in AI yet as well to make it a truly useful tool that isn't a gimmick.
Michael: At the annual HPA Tech Retreat a few months ago, a multi-day event, at almost every session, every discussion, MovieLabs was discussed. Not just discussed, it actually had a place of prominence on the screen. HPA, a lot of times, is very much feature film, larger budget broadcaster, cable television oriented. And I would love for folks who are more independent to kind of know where MovieLabs fits in for them when they're not working on those types of shows.
Mark: You know, MovieLabs is owned by the studios. They produce a huge amount of episodic TV, right? Just to be clear, they are not just making movies. So episodic TV has always been included. I think one of the reasons why you saw it mentioned a lot, but it's not MovieLabs that's mentioned. It's the Vision that's mentioned. The Vision has been democratized, and there are a lot of companies that now have it as their vision, right? It's their strategy. It's, "This is where we are going, and the studio said they wanted to get there." Hell, that's what we wanna go to too. So it comes up a lot because I think a lot of people, it landed at a time when a lot of people were going, they needed direction. They'd heard about cloud, they'd send some bits of this, and it just put everything together on a nice little packet.
People went, "That's it. That's what we've been talking about. That thing." If you look at those principles, you could apply those to a webcast at a conference. You could apply those principles to making a student film, making a 30-second commercial, you know, making a YouTube video. I mean, they are pretty foundational for all types of media. And actually, there's some cloud companies that have been talking about them. They're pretty useful in any creative industry, actually. They're pretty good high-level principles for everybody to work for. We may create tools for the very high-end that can afford, you know, 25-million dollar visual effects budgets and given builds of amazing virtual production technologies. But if we do it right and we create enough scale, those innovations will fall down to everyone, and you will be able to do, you know, virtual production on an iPhone.
We're seeing begin bits of it now, right? You can start doing swapping out backgrounds and stuff. Even five years ago, that was unheard of. You know, you can get a visual effects company to go and do rotoscoping now, and we're getting to the point where some of this technology is becoming democratized as well. So I think the Vision was published on behalf of the major studios, but it is not owned by them. And it is a vision for the future of creative industries in general. And I think that's the way it should be perceived.
Michael: Lastly, where are the various places online where we can read the paper, read the updates, see video fireside chats, etcetera? Where can people learn more?
Mark: So MovieLabs.com is the best place cause it's got the showcases on there. There are some video things that we put out as training. There are more of those coming up. The visual language, which we haven't spoken about actually, but we have a whole visual language. We're building workflows. All of that is on there. The ontologies are on there. And then the best way to find out updates is to follow us on Twitter or LinkedIn, and you can look us up on MovieLabs on LinkedIn, and we're gonna start doing a newsletter soon as well. Because there's so much stuff going on that sometimes we don't even hear about, and we go, "Whoa, whoa, that happened?" So we're gonna start putting together a newsletter. It just brings it all together so people can get one digest of all the things that are happening. But yeah, there's a lot to follow. It's going well.
Michael: Thank you so much for your time.
Mark: Thank you.
Michael: Thank you for tuning in. This has been Mark Turner from MovieLabs. I'm Michael with Shift Media.
Miss our interview with Terri Davies, President of Trusted Partner Network? Watch here to learn more about the TPN application process.
For tips on post-production, check out MediaSilo’s guide to Post Production Workflows.
MediaSilo allows for easy management of your media files, seamless collaboration for critical feedback and out of the box synchronization with your timeline for efficient changes. See how MediaSilo is powering modern post production workflows with a 14-day free trial.