How Cloud Tech Is Changing the Way We Make Movies with Hugh Calveley

On Production with Hugh Calveley

‍

Welcome back to On Production, the Wrapbook podcast that pulls back the curtain on what it really takes to bring stories to life in film and television. I'm your host, Cameron Woodward. Today's guest is Hugh Calveley, filmmaker turned entrepreneur and co-founder of Moxion, the New Zealand based platform that transformed how production teams review and collaborate on footage. What began as a common onset frustration grew into a tool trusted by Warner Bros, Netflix, and Amazon turning production dailies into immediate secure, shareable content across continents. In 2022, Autodesk acquired Moxion. Today, it lives on as a part of Autodesk FlowCapture, serving productions globally with cloud-based workflows. But my hope for this conversation isn't really just about tech. I'm also very curious about how Hugh thinks about building something indispensable, navigating Hollywood from New Zealand on the edge of the Pacific and discovering how constraints can spark clarity. Hugh, thanks for joining me and welcome to the podcast.

‍

Hugh

My pleasure. I'm so excited to be here.

‍

Cameron

So I was able to do some research on your background and I realized that you worked on set as a camera assistant early on in your career. What did those first roles teach you about hierarchy, information flow, urgency, and how did that shape your approach when building Moxion?

‍

Hugh

I always say that joining the film industry was like running away and joining the circus because you have such a eclectic bunch of people working on the film set. And also, there's this wonderful generosity of spirit that goes on too. Because no one wakes up in the morning and says, you know, I want to make money, so I'm going to make movies, right? Everyone's there because they just love storytelling. yeah, I I had a background in computer science and I had kind of always actually been interested in movies. And I was reflecting on this the other day and I think if I'll just divert a little bit. But I remember when I was a kid, I grew up in the country, I was probably five or six years old.

‍

On a gravel road, way out. The nearest grocery store was like a day's walk kind of scene. And I do remember that one day a film came to shoot next door and it just blew my mind because there these lights, I remember shining through the trees and it was like the aliens had landed. And so of course I had to go and look and I was pretty feral. So I remember the lights and I remember lots of shouting, things like, get that kid off the set and there's that kid again. But looking back, I think I was kind of hooked from way back, which when I joined the film industry was at a really interesting time because it was at the point where the digitization was just starting to be hinted at. So the offset, the post had probably been digitized, you know, five, 10 years earlier and

‍

I was on set when I was actually doing this crazy thing when you're filming using film, right? So, you you, talk to the new folk these days and they kind of go film. Did you touch film? You know, what, did it feel like? Right. And the digitization happened so quickly that it felt like one day I was in and changing mags. And then the next day I was swapping out cards. I think reacting to that and seeing all the silos that sprung up in terms of all the different data, that really kind of, I think, clicked in my mind. went, we've got to get all this precious metadata that is being kept on set or sometimes, you destroyed on set into the hands of the post people that really needed to do their work. So I think that was probably my first experience was this new world and this incredible transition to a new state of play. And then responding to how do we make this easier for people to tell their stories was really kind of the inception for what I did next.

‍

Cameron

‍

When you say metadata and sort of feeling like there was this sort of sadness related to it, just sort of going into oblivion instead of being used, can you describe what bits and pieces of metadata you felt like were particularly valuable that were being wasted?

‍

Hugh

Yeah, sure. So I would try and think of some real practical examples. lot of the time, so simple things like scene and take and all that kind of stuff, but a lot of the stuff from the camera in particular. So what lens they were using, what was the stop? So what was the focal distance there? And if that information didn't travel with the dailies, then either it would have to be

recreated or guessed at further down the line, especially if visual effects were involved. And that would take a long time and it might not be accurate as well. So some of that stuff is just so key to what happens further on that. unless you capture that, it's just going to be an impediment to the whole filmmaking process.

‍

Cameron

That makes a lot of sense. So in past interviews, you had articulated that Moxion was born out of frustration. We were just discussing metadata, but was there for you like a last straw moment when you thought somebody has to fix this and maybe it's me?

‍

Hugh

Yeah, yeah. I I've still told the story where I was, I was sitting in a car and I had these camera sheets in my hand, which kind of are critical metadata. And I had, I'd been working on this, this crazy horror movie and had lots of vampires and I'd been given the camera sheets, which would indicate amongst other things, whether the film had been pushed or not.

‍

I was trying to get them to the right people and I got pulled over by a cop and I was just freaking out because I had driven a little bit, how should we say, spirited. And yeah, I was sitting in this car going, my whole film career is over because this cop's just taking his sweet time and I've got these camera sheets with this vital, vital metadata for the film. And I think back then I was going, you know what?

‍

that camera knows that metadata. If we could somehow connect it to this new thing called the cloud, then I wouldn't be sitting here worrying about my future employment. So I think that was kind of the impetus. And then from there, things evolved. The cameras became more digitized. More of the metadata became available. And I think the second light bulb probably went off when I was working with my co-founder, Michael, on a shoot.

‍

And he was the editor and I was doing the dailies and he made a comment to me. He said something along the lines of, if I got this footage yesterday, I could have told them they didn't need to shoot half of it. It's not going to be in the edit. and that's when I went, we can probably do something about this, this now. and we, we came up with a means of essentially what they now call camera to cloud, but back then we called it immediate, of getting the footage directly from the camera into the cloud and into the hands of the editors or anyone else in post that needed to get a leg up on what was happening.

‍

Cameron

That is a framework that I had not considered before. When you start thinking about selling software into a studio, you can sell efficiency, but describing it as Michael did, like, actually, we can probably cut time out of a production schedule because this is unusable or unworkable anyway. And just making the entire workflow more efficient, like that's a really interesting insight that I had not considered until just now. That's really, really interesting.

‍

Hugh

Yeah, and it also gets interesting when you start to, I guess, let additional units see what the other units are shooting in real time too, because you would quite often shoot to the storyboards. But if you actually see that, the main unit is shooting a much tighter shot than what was storyboarded, you can, as a second unit that's following the main unit, go, you know what, we don't need to shoot

‍

such a wide shot, which has implications on costs. Suddenly the set people don't need to build out that wider set. You can dress it right down. You can cheat it a lot more. And it just makes the whole process so efficient. mean, mean, and that was one of the first things, you know, that I, I saw coming to a, a film environment, which is, I guess a little bit like a startup too. You get all these creative people that come together for one time, and their goal was to tell the story and they start to figure it out and by the end of the shoot, they figured it out. But then they all go and they do the whole process again. It's kind of really inefficient. Like they get the big things right, but I remember as a runner, one of my first jobs was to drop off these salami sandwiches to a homeless shelter because on that day they'd canceled the extras, but they'd forgotten to cancel the food. So I've always thought that there are probably better ways of controlling the change that inevitably happens when you start shooting.

‍

Cameron

You're, you're sort of entering into my world of production, finance and accounting and production scheduling, et cetera. But it is amazing how even, something like dailies technology, workflow technology that leads into post-production. There's implications there all the way during production as well. So it really is amazing how interconnected these workflows are, even if there's radically different departments involved. were touching on this a little bit, Hugh, which I think is interesting. So early on, I'm curious.

‍

What did you misunderstand about the problem you wanted to solve? And in hindsight, where you were wrong, but also maybe what did you surprisingly get right?

‍

Hugh

I think the biggest misunderstanding was the just because there's a better way of doing something, it doesn't mean people are to drop and adopt it. Right. It's you know, we know what we like and we like what we know. And I think that you really need to get the right production and the right people at the right time. Timing is key, right? So you have to, and also I think, yeah, you just have to have to be in the right, there's a lot of luck. There's a lot of luck. yeah, the adoption that people were just going to adopt it and that was tough for us too. So as a startup, we got fairly quick adoption of what we call the mediates inside New Zealand and into a little bit in Australia. And I think that's because things were like a little bit more, I don't know, rock and roll down here. you have tighter budgets and you have to figure out ways to make it more efficient. And so we went to to Hollywood going, we're to change the world. And people were much more considered up there. They would go, yeah, yeah.

‍

Well, we'll think about this a little bit. so we didn't really get a break until this wonderful man called Barry Osborne, who was the producer of Lord of the Rings and The Matrix. He took a real gamble on us for a couple of movies. One was The Meg and the other one was Mulan and could really see how this could help some of his problems. Like, for example, on The Meg, he had this need for a press boat to be near the hero boat, but far enough away that it wouldn't be in shot. And he wanted them to see what was being shot on that, of that hero boat. And they were going to do this crazy microwave bouncing around to try and get it back. And it was going to cost them an absolute bomb. And I said, well, we can do this for you for the cost of like a 4G connection, right? Which was probably 3G back then. And he said, yeah, can you do it? And I said, yeah, absolutely. And he said, well, you better because you don't want to get this wrong. And so we spent the next month building what we said we could do. And it just worked wonderfully they were hearing cut on their earpiece in the press boat and 90 seconds later they were seeing on the screen what had just been shot. So that had gone all the way from New Zealand up to the west coast of the US and back down again in about 90 seconds.

‍

Cameron

That's incredible. Maybe one day you'll get into the business of transferring data from Mars to Earth. Who knows? It's so interesting, It's surprising to me that it was New Zealand where this technology was born and raised and then became really a standard in many cases across the entirety of the industry globally. New Zealand seems to punch above its weight in film technology. know, famously, course, Weta is there. What cultural or industry factors contribute to that innovation, you think?

‍

Hugh

I think it's a combination of, there is a bit of a pioneering spirit, right? Which is still there. And we just don't have a lot of resources. We just have to make do with what we've got. And I think that leads to some fairly lateral ways of solving problems. And yeah, on a film set, you just have to solve the problems. I I love that quote. I'm going to say his name really wrong. It's Alejandro Inarritu. But he says, you know, to make a film is easy, to make a great film is war. And under the pressures of war, you can just see how creative you can be. And I think that comes out.

‍

Cameron

Totally, really high stakes, not a of resources, but we must get to the end and see what happens.

‍

Hugh

Yeah, yeah. And it's probably the same in the US. I haven't been on a lot of shoots in the US, but in New Zealand, there's certainly a really eclectic bunch of people. you get like physicists walking, working next to gang members sometimes, right? Or ex-gang members, I hope. But just that melting pot of people who can physically build things and who can then also mentally build things is a really interesting combination.

‍

Cameron

So I'm really curious from like, you were just sort of describing that spirit of ingenuity, the Kiwi way, I suppose, but from a computer science product perspective, actually building this product to have the outcome that you did on that film. I mean, you were doing something really at the cutting edge of technology. As you develop Moxion, what sort of design principles did you prioritize? Maybe ones that other technologists were overlooking?

‍

Hugh

Yeah, yeah. Well, I think, I mean, this is where having Michael as my co-founder helped because he had a background in design as well as being an editor. I mean, it's always really, I mean, guess the two answers that is, you know, what do you build and what's the interface with what you build, right? The design, right? So, and that's always like a tightrope walk between building what people want, what they're asking for and what they don't know they want, right? Which is the real innovation, right? So, you know, I guess, do you build the, do you copy the existing cell phones or do you go build an iPhone, right? And that's really hard. But I think, you know, in terms of design, it's also, it's a tough audience, right? They're all creatives. So it's got to look not just good, but the right kind of good. So it can't look like a toy, but it can't be too complex, right? And my co-founder Michael would always say, what we've got to build, the criteria is it must delight them. And that's what I think we strive for.

‍

Cameron

To that end, was there ever a feature you considered but ultimately resisted? Did you decide what not to build when DPs, execs, and producers were all pulling you in different directions?

‍

Hugh

I don't know about that, but there is certainly a feature we have built which we were under pressure. We built it. We didn't adhere to it, Master Light them. And I'm smiling because it still haunts us. And our users will know what it is. And we're doing our absolute best to remediate that.

‍

Cameron

Of course, I'm in software too. Hugh, if you had to explain what makes FlowCapture beautiful from an engineering or product standpoint, what stands out?

‍

Hugh 

I think like, well, the very early adoption of what the cloud can offer from an engineering point of view is pretty fundamental. for example, very early on we adopted microservices and that, you know, is really the ability to run lots of little bits of compute in parallels. So for example, to answer your question, making that decision early on enabled us to do some really clever things. this whole process is really, speed is really important, right? So we have a saying that nothing kills creativity like having to wait. So if you're uploading something, it's super important that that is processed and made available as quickly as possible.

‍

the, also the other thing, and we haven't been touched on it, but one of the key principles we always adhere to, and I think my computer science background helped me here is that there's so much time and sweat and tears goes into this, that it's gotta be whatever you're sharing has to be secure. Security is absolutely paramount because all that toil and effort is wasted if it's not secure. So in a very, very roundabout way, I'm going to answer your question. and.

‍

What? Engineering it correctly with microservices enables us to do is to process the assets very very quickly we can split say a video up into 10,000 bits and have it processed through 10,000 little microservices in parallel at the same time and When we do that we do a whole lot of things we are creating probably about six or eight different versions do you want to see it in in Dolby vision or do you want to see it on your cell phone and marginal reception with 480p? Right Do you want to have forensic watermarking built burnt in so like invisible watermarking? Do you want to have visible work watermarking burnt in all with your personal details all in real time? And the actual compute to do that is pretty pretty massive. So I think it's the unsung hero of our platform is

‍

Cameron

Did GPUs make that possible? What was the parallel computing pro, like how did.

‍

Hugh

no, just the well, I guess two things, right? The idea of microservices so you can you can virtualize these little machines, you know, and have them have them work for you. And but also just the ability to expand and contract very dynamically using cloud resources. yeah, it's it's it's it's amazing. So you you you play a video and in real time you are having burn in identification. You're having forensic done and also digital rights management, you know, the ability not to be able to download it or take screenshots. That all happens pretty much in real time.

‍

Cameron

That's incredibly cool. This is a bit of a pivot a little bit, but I'm just curious from your experience. Is there a persistent myth producers hold about digital workflows? Maybe not even at this point anymore, but that you'd just like to retire immediately?

‍

Hugh

Yeah, I've got one. I've got one. That they are inherently unsecure, right? In terms of you're putting all your eggs in one basket. That you shouldn't hold everything in the cloud in one place because it just makes everything open to be gotten from one place. I think the analogy I always use there is, yeah, would you rather have it in a multitude of places and each one of those places has a very weak padlock, right, with an easily picked lock? Or would you rather have everything guarded with the most expensive technology known to humans, know, with ninjas with lasers for eyes, you know, protecting you. And so I think, you know, the cloud is safe and it's much safer to have it up there with, you know, well known, well tested security than to try and squirrel it away in a multitude of places.

‍

Cameron

I like that one as well. You know, I deal with a lot of payroll data with our business at Wrapbook and you know, the state of the art seven years ago was like people's personal information in banker's boxes sitting on a desk. And it's like, actually the digital storage vault is quite more secure than that. You know, that makes sense. So you built this business, you had some lucky breaks, you built some really beautiful novel technology that solved this very painful problem. produced a lot of like, you know, in the course of production workflow upside, but then also even going in towards post-production, right? Based off of everything that you've told me. But in 2022, Autodesk, which is a huge corporation, very well known, you you think about like AutoCAD or like, you know, people listening to this, they know what Autodesk is, right? That was a landmark.

situation, what did that do for you and the product? What did the product gain?

‍

Hugh

Yeah, yeah, really interesting question. I mean, I think, I mean, when you're running a startup, like I was, you wear like a multitude of hats. But you're not that conscious of them, right? You're just getting the job done. And I think, in a way, even though I've moved to Autodesk, those hats are still there. I just know the names of them, right? And they're kind of entire ecosystems. you know, like technical sales, support, marketing, customer success, etc. And I'm still involved with all those departments, but I'm not running them anymore. But I think that what has really changed is just the ability to scale, right? Because I'm not having to run those departments. And there are people who are far more capable than I ever was in charge of those departments.

just gives me this ability to realize what was in my head faster. yeah, yeah, it's just, I mean, to take a step back, one of the reasons we went to Autodesk was that they kind of had a vision that we shared, but we knew we could never do it, but they can.

‍

So our vision, we even had a name for it because when you're a startup, you've got to have like a North Star, right? If everything goes right and we raise a gazillion dollars, we're going to be there. And this was kind of, we called it the fabric of filmmaking and they had a very similar vision. It was more developed and, but they called it Flow. And this was really the concept of a production platform in a cloud. In a shared environment where you can use as much or as little as you want, but there is a commonality of data that you can access, metadata. I think, this is going to be even more important with the storm is coming and the storm is AI, because AI is going to be able to leverage everything that you expose at a platform level. And so the ability to have a common API or common APIs that are exposed is just going to be absolute fodder for AI in a great way.

‍

Cameron

you mentioned earlier at the beginning of our interview, all of that precious metadata, that's still true to this day, especially in the age of the wave, the tsunami wave of AI that's coming in filmmaking. It should be super interesting.

‍

Hugh

Yeah, I mean, I've been lucky enough to be around to complete upheaval. You know, the first one was the digitization of the onset and I see AI as being the second wave of radical transformation, probably way beyond what happened on the film set.

‍

Cameron

Yeah, mean to that, do you feel like you're solving the same problem or has the landscape and priorities shifted?

‍

Hugh

wow. I think it's shifted. mean, the same problem is always there, right? It's really about getting Watson in here out there as quickly as possible. But AI is going to radically transform that. And whole workflows will be disrupted. Things that happened in sequence will happen in parallel.

You'll just have this incredible ability to just very, quickly test ideas and not have a huge cost to test, think is a really big one. But that's so expensive now. Yeah.

‍

Cameron

It's super fascinating to that end. I'm really curious in your view, what's the most under discussed constraint in production workflows today? Something invisible, maybe that if solved would, would change everything.

‍

Hugh

The ability to order a coffee on an app on set. think that would radically change the way we make movies. So I think some really interesting stuff around data analysis, I'll take a step back. Everything starts with a script, right? And we can go into what AI would have a script making, but just assume that, you know, a script, script comes in and you want to make this. if you look at, if you can break down the elements of that script and see what you might already have, then I think this will be really interesting. So if you look at some of these big film studios, they have this huge amount of archive material. And they don't necessarily know what it is. So I've been on countless movies where we've gone and shot muzzle flashes or breath elements again and again and again. There must be thousands of these breath elements or muzzle flashes. And they will be, but they don't know how to get them. It's cheaper to go and reshoot them than to go and find them. And I think that's about to change radically because as soon as you let AI loose on their libraries of information, right? And there'll be so much stuff which never made it to the final cut, all the B-rolls and all this kind of stuff, you're gonna have this treasure trove of footage, which can probably be applied to a lot of scripts. You'll put the script in and you'll go, okay, we need breath elements. Well, tick, tick, done. New York skyline, done. And Yeah, I think just that ability to quickly, quickly break down a script and see what you actually have to go and shoot or what you actually have is going to change everything.

‍

Cameron

Hugh, that's a really compelling and interesting vision, and I think it plays really well into Flow and what you're building towards with Autodesk. It's really awesome. I'm curious, totally different thing. If you could wave a wand and change one global production norm or habit, what would it be? And what would be the unintended consequence?

‍

Hugh

If I could change a production norm, it would be involving the editor a lot, lot sooner than they currently are. And it's a hard ask because editors by their nature, know, are creatures of comfort. They don't like to leave their nice warm little editorial cave and go on set where, you know, there are bright lights and the coffee isn't very good. And I think that if we could involve them through technology, then we could really, really get a leg up because what happens is... And actually we started to do this. We started to do this on Rings of Power. So it was a really nice circular arrangement where the editors got the footage quickly from set and then they would very quickly edit together in a rough cut the scene that the directors were working on on set. So I think it was like within half an hour, the directors could see back on set what they had shot half an hour prior in the context of the scene that was being edited. that transformed how some of them actually worked. They would go, you know what, we've got this. We're gonna move on. We're gonna focus on the stuff that really matters. Getting the actors in the moment. We don't need to shoot anymore of this coverage. It's not going to be in the edit.

‍

Cameron

So helpful. Over the next 10 years, I mean, you were speaking to it a little bit, but just based off of like being able to truly have like what Vannevar Bush described as like a huge hypertext library, but in this case, it'll be with footage. So if I need a skyline, I can find a skyline. It's great. I can license it, whatever. But over the next 10 years, how do you see the role of the producer evolving?

‍

Will they need to be deeply technical, do you think, or will tools just make it invisible?

‍

Hugh

I'm hoping it does make it invisible to a sense. mean, I think there going to be some wonderful, I just think about this the other day. We always talk about like the metadata and the data informing what we shoot and everything. But I was thinking it might also give us some really interesting insights into who's doing the shooting. So as a producer, you can probably put together different teams and get some really good metrics on how they're to perform together. So for example, you you might find through analysis of past productions and budget that if you put this DP with this director, they're to shoot some crazy ratio of like, I don't know, a thousand to one, right? And that's going to blow your budget. So I think there'll be some really, that they'll be, you know, what's, what's the thing that you can predict anything about the future?

But I'm going to go out on a limb here. And I think they'll become more and more kind of like data analysts than actually the current roles.

‍

Cameron

We're seeing a lot of that in the industry, even over the last number of years, but as like machine learning just gets better and there's more and more compute, you could imagine that our ability to predict these outcomes will get better, which is really, really fascinating. It's an interesting thing. I have one final question for you, Hugh, which is, what's a question you wish people would ask you about your work, but rarely do?

‍

Hugh

yeah. Why do I do it? I think I think is a question that I don't think I've ever been asked. I guess I got to answer that now too, Yeah.

‍

Cameron

You have to answer it.

‍

Hugh

I do it because there's nothing else that gives me the kick I get from when someone comes up to me. And this was actually real. Here's a real story. was, fairly early days. I was in a studio in LA and, I was about to leave. was walking out and this large bear-like man jumped out of a room and said, are you Hugh?

‍

And I went, oh, but I was at the studio and had like a name badge that's big on my chest. So there's no point denying it. Right. I said, yes, that's me. And he said, oh, I freaking love you, man. And gave me this big hug. And what had happened was that he was in the visual effects supervisor or he was in the visual effects team was that the immediate we were doing had all the visual enough metadata so that they could finish the visual effects budget by the end of each day, something they'd never been able to do before. And he could go home and see his kids. So I think that kind of response is what gets me up in the morning.

‍

Cameron

That's awesome. Hugh, well, thank you so much for sharing your story. Thank you for building really awesome technology that makes the production process work. Really a pleasure to meet you and really excited to see what you do with the entire suite of tools at Flowcapture.

‍

Hugh

Thanks so much. My absolute pleasure being here.