Proceduralism in the Metaverse

Announcer:

Today on Building the Open Metaverse.

Kim Davidson:

Photo real effort is really, really important to the customers that we're going to, but if we can do stuff faster in performance, give them some feedback on that explosion, just even a piece of that explosion, or a low level, a low thing in real or near time, the faster they can get to the photo real look.

Announcer:

Welcome to Building the Open Metaverse, where technology experts discuss how the community is building the open metaverse together. Hosted by Patrick Cozzi from Cesium and Marc Petit from Epic Games.

Marc Petit:

Hello everybody and welcome to our show, Building the Open Metaverse, the podcast where technologists share their insight on how the community is building the metaverse together. Hello, my name is Marc Petit from Epic Games and my co-host is Patrick Cozzi from Cesium.

Marc Petit:

Patrick, how are you?

Patrick Cozzi:

Hi, Marc. Doing great. Very happy to be back.

Marc Petit:

Yeah. Are you looking forward to today's episode?

Patrick Cozzi:

Oh, I can't wait to talk procedural.

Marc Petit:

Yeah, today we are going to be talking about content creation in the metaverse. And we invited someone who is an expert in the domain. I don't know if we're going to call you a pioneer or a veteran, but please welcome Kim Davidson, the Co-founder, President, and CEO of SideFX software.

Marc Petit:

Kim, welcome to the show.

Kim Davidson:

Oh, thank you. Great to be here. Thank both of you for having me on.

Marc Petit:

Yeah, it's a pleasure. So you're a mathematician by trade, from University of Waterloo. And you founded your company more than 35 years ago. Who knew the industry? Patrick, you didn't know the industry was that old, did you? [laughs]

Marc Petit:

Yeah. So in 1987, you had the goal at the time of bringing 3D graphics to a wider audience. So that was an ambitious goal back in '87. It was originally known as Prism and the software became Houdini in 1996. So during this journey, you and your colleagues at SideFX collected five technical academy awards, including one Oscar, the famous statuette, so congratulations for that.

Marc Petit:

Please walk us through the creation of your company with your partner, Greg Hermanovic back in '87, because that was the early days, right?

Kim Davidson:

Yeah, I'd sure love to. This could take the whole hour, but I'm trying to be brief. Because I love history because it gives us a window into the future, which is what the metaverse is really something that we're all aiming for at the moment.

Kim Davidson:

So Greg, I'd have to go back to, say 1983, where I first met Greg. And we both worked at a company in Toronto called HCR. And they had a successful growing business, porting Unix operating system to mainframe computers and minis. Greg and I did work on some computer graphics projects, but that was very early days.

Kim Davidson:

But I'd say our real start, guys, was just prior to SideFX, was when we left HCR and joined Omnibus Computer Graphics in Toronto in 1985. And to give you an idea, at Omnibus, that was the days of doing flying TV network logos, commercials.

Kim Davidson:

And there were only a handful of companies. They were about 75 people over three locations, Toronto, Los Angeles, New York. They did work some work on Flight of the Navigator, so there was a little film work. That was done out of LA. In Toronto, we were doing the flying logos.

Kim Davidson:

So these are definitely the early days of computer graphics for the media and entertainment industry. And we're working on mainframe computers, tech terminals, one frame buffer, recording on one inch mag tape. This is all working blind. But it was a lot of fun.

Kim Davidson:

But there's no commercial software. This was out of universities where we were getting some code and some software out of the military. So we're basically, Greg and I and another team were there to write the software. And it was the president of Omnibus, John Penny, who actually named the software for one of the... They were a public company so they had a stock exchange.

Kim Davidson:

For one of the annual reports, he named it PRISMS. And it was actually an acronym, which I like to say. And I wrote it down because it makes little sense. It should have procedural in there in my mind, but it's the Production of Realistic Image Scene Mathematical Simulation. And that's what PRISMS was.

Kim Davidson:

But there was no program called PRISMS. It was just really a collection of Unix programs that you could put together in a script. So we'll talk about, it was really the roots of proceduralism right there, in putting those commands together. But we'll talk about that together later. But basically string these commands together, give them a bunch of arguments. And if the client wanted his logo bigger, thicker, more bevel, you just change a couple arguments, rerun the script, and out comes the geometry and out goes the renders and all that.

Kim Davidson:

So we're kind of a pretty, I don't know, I don't even think it was clever, it was just the Unix and Seashell way of doing work, leveraging the computer to its advantages.

Kim Davidson:

But in 1987, after Omnibus bought Digital Productions, which worked on The Last Starfighter and Robert Able & Associates in LA, they went Omni-bust. They were, as some people like to say, Dead On Arrival, Digital Omnibus Able, DOA. But at that moment, for maybe a couple months, they were the largest company in the world doing this commercial animation.

Kim Davidson:

So in 1987, Greg and I are on the street, but we decided to put a bid into the receivers and bankruptcy for the exclusive right to the PRISMS code. We got PRISMS, we incorporated SideFX, and we just picked up where we left off, just the two of us. So albeit a much, much smaller company, but we knew how to do animations. I was the animation director.

Kim Davidson:

But I'd say, guys, beyond the code, really those four years of experience that Greg and I gained at HCR and Omnibus in the industry, and using the software and writing it, that was the invaluable part of starting SideFX.

Patrick Cozzi:

Yeah, that's a very unique founder's journey. Yeah, congrats on the success and what you've built today. So Kim, usually for the podcast, I'm the geek and Marc is the responsible adult. So I like to ask some of the geek and tech questions. So to start things off, Houdini is super widely used for visual effects. And some of the best visual effects I've ever seen have been made with Houdini. I was wondering if you could tell us a bit about the tech and the procedural tech that's used?

Kim Davidson:

Well the tech is C++ code, but I don't think that's what you're asking. But if you sort of mean architecture, then Houdini's like this flexible, full 3D animation system. And it's based on a procedural kind of architecture. That means everything from modeling to animation effects and compositing can be changed, the way I was talking about just that simple example of the type earlier.

Kim Davidson:

And the other thing that's really, really key in the software, it didn't start as a CAD software. It started with time. Like tying the fourth dimension is very integral to the design. Anything that you want to animate can be animated. So they always had this procedural approach, PRISMS and Houdini. So it's really built into the core of the system.

Kim Davidson:

And I think, because we were animators in the early days, even when we started SideFX, I animated probably over a hundred spots in the first year, just paying back the loans on buying the software.

Kim Davidson:

But because we're animators and programmers, I think we're eating our own dog food, I feel like. So we create an architecture, continue that architecture that we could change, that the directors, women and request. So maybe I like to think programmers are lazy, they don't want to do anything twice.

Kim Davidson:

But it's certainly efficient, and if you don't want to start from scratch. So quick changes, experiments as well, giving you a lot of varieties and their abilities for a director to pick from, if you want to show them to them at all. Because often, you don't want to give them too much choice either, but you have them in your back pocket.

Kim Davidson:

So anyway, I do want to say though, it is a full 3D animation, but we're known a lot for the visual effects. And there's a reason for that, is because there was a lot of big companies in the mid 90s, as graphics cards were taking off, and companies were getting absorbed and bought. So we went to the one segment where we could really provide the biggest value with this proceduralism. And that was the visual effects, and in those days it was film, in Hollywood primarily.

Kim Davidson:

Because visual effects then and now, is where you really want to iterate to see what it's going to look like. It's a simulation, you have to sit back and wait and see what that's going to look like. So that's where the proceduralism really kicks in.

Patrick Cozzi:

And everything you're saying about the dogfooding and the fast iteration time, same way we write code as well. So it's great. And speaking of the visual effects, what's next? What's on the roadmap?

Kim Davidson:

Well that's the cool thing about this industry, Patrick, as you know, there's always the next big thing, and there always will be sort of thing, it always has been.

Kim Davidson:

The one thing we're trying to do, as we grow into these different segments and different parts of the pipeline, is just make the VFX, the visual effects, more accessible. So that's a lot of prebuilt things, bringing more usability, more smarts to the whole program. So that's really one thing.

Kim Davidson:

The other thing is really leveraging our knowledge in effects for doing more simulation for characters and character animators and artists. Maybe slightly further out, it's not a new topic, it's its own podcast, is machine learning. We've done some experiments on our customers, but Houdini can push a lot of data, create a lot of data.

So as an example, you wouldn't want to tweak a few parameters and run the explosion, because it's like, oh, that doesn't look right. You have to re-tweak them. Why not let machine learning do thousands of iterations and see what that looks like? Because, basically, it can match a piece of video footage of an explosion that's already to the director's liking.

Kim Davidson:

I think the other thing, that I'll mention one more thing that's next for everyone, and certainly for Houdini and visual effects, is just more and more real-time. We've been pushing... Photo real is really, really important to the customers that we're going to, but if we can do stuff faster and performance, give them some feedback on say that explosion, just even a piece of that explosion, or a low level, and real more near time, the faster they can get to the photo real look.

Kim Davidson:

And the interesting thing is, the game developers are coming from, they start with real-time. And over the years, as you've seen with the engines, photo real, photo real. So I think working with a game guy, for sure guys, is really helping bring the best of those two worlds together. We're learning a lot from working with them, and they're, I think, upping their game on the photo real as well.

Marc Petit:

No, I was fascinated by the prospect of using machine learning to augment artistry, to really support the artist. I'm glad to hear that you're doing experimentation. So are you optimistic that this can deliver on its promise of really taking a lot of the duration of the cycle?

Kim Davidson:

Oh, absolutely. We've done a test on machine learning, sorry to cut you off there, Marc. Just paint, and you have Rocky Mountains. You paint and then use the machine learning, just takes the Rocky Mountains and builds the Rocky Mountains because it already knows what red is and blue is, and creates a realistic landscape.

Kim Davidson:

It's really quite cool. On Into the Spiderverse, I think we have a case study on the website, they used a lot of machine learning for the ink lines. It was Houdini, but they set up the whole system and ran it through Houdini so the artist didn't have to paint as many ink lines on the CG characters to make it look a little more cartoony.

Kim Davidson:

The sky's the limit with that machine learning. And AI, obviously beyond that, that's to me a whole other level of intelligence. A lot of people say AI, they really mean machine learning, I think, in the two use cases I just talked about.

Marc Petit:

And even neural networks. We don't want to be overly technical, but we've seen this thing fall out of fashion, neural networks, and come back very, very rapidly in the concentration process.

Marc Petit:

And it's very relevant, because with the metaverse around the corner, the demand for 3D content, and original 3D content will explode. So I'm sure these architectures will help people like me with no artistic skills to actually create some of that content.

Patrick Cozzi:

And Kim, do you think that procedural will reach every creator? Because we're seeing everyone potentially becoming a creator with the metaverse.

Kim Davidson:

Well, and I think under the covers in some parts, you want to speed something up, you make that part procedural and that part hard-coded or locked in, and don't give a choice, because that's not needed, or not necessary, or not for everyone. So definitely, having a lot of stuff done on the fly can be very, very powerful. Call it procedural, call it routines, scripted, whatever you want, but we kind of coined it procedural because it is a recipe.

Marc Petit:

So let's come back on real-time a little bit, because we care about real-time. We care about the metaverse. And I remember in 2013, you remember, I know we mentioned Toronto, and you had launched the Houdini engine. And it was a little bit of a surprise to me that Houdini and games is a thought that actually had not crossed my mind back in the time.

Marc Petit:

But since then, you really proved the value of your technology in the game development process in the real time 3D. And I think we can say that Houdini has been used on some major Triple A games now to help the process of creating those open worlds.

Marc Petit:

Can you talk to us more about gaming, and Houdini, and how the proceduralism is helping those guys as well?

Kim Davidson:

Yeah, thanks. That's great memory, Marc. After the film visual effects, we did say we'd always had some customers in games. I think our second customer was in games in Japan. But after the visual effects, we consciously decided.

Kim Davidson:

But we always had this Houdini engine. It was used to run anything that was compute intensive in parallel in the background. Anything you could type in Houdini, you could make and run in a command in background. So it's always been there.

Kim Davidson:

But I think in 2013 was where we were really starting to think about APIs, and plugging it into other software, and creating that connection. And that really worked for, especially for the gaming people who didn't want to leave their package and jump back and forth, save something out into an Alembic file, go back in, oh, that's not quite right.

Kim Davidson:

They just wanted to stay in the editor. So with Houdini and these, what we call Houdini assets, smart Houdini assets under the covers, you have the little interface presented that can create a tree or a bunch of trees, and a bridge and a bunch of bridges. And they only have three or four controls, but it's really Houdini doing the re-cooking whenever you change one of those controls under the covers.

Kim Davidson:

So that's the first place. And the first place in games specifically, was environments, because those are bigger problems than building a box or a car maybe, where you can buy a car and modify it, or build one car and it never changes up to a spec.

Kim Davidson:

But environment's kind of something that just keeps going and growing. It's very organic. So that's where the Houdini engine’s, a lot of tools, started building there. And so to give the listeners just a quick example, what does that really mean?

Kim Davidson:

I'm sitting in the game editor, I'm building it and I want to change a road 50 feet to the left. So you just move the curve that road is based on. You don't pick up all the geometry. You move the curve, the road gets recalculated, geometry of the road, the geometry for the guardrails, the geometry for the lights that's on all the rocks that are around.

Kim Davidson:

And literally, if there was a mountain in the way, you'd just move the mountain, literally if necessary. And then it's like, oh, let's gameplay that. Does that work for you now, Mr. Level designer, or gameplay designer? I'm showing my growth in the wrong industry. I'm trying to play games as much as I can, Marc, to keep in that space.

Marc Petit:

So you're the guy who moves mountains easily?

Kim Davidson:

Yeah. Yeah. And further, just to make it a little easier, we did create something called Houdini Labs about four years ago. And it's there so we could just iterate, so little tools to fix UVS or to poly reduce. But poly reducing isn't one command, it's like five little commands that you want to tweak. But you just want to expose one or two things.

Kim Davidson:

So we created Labs and we contribute to that, and we invite other people to contribute to Lab. They're on Github, but also on our website where you can download these little tools that work with Houdini.

Marc Petit:

It makes sense. So you're making the process of creating those worlds much easier and almost fun, actually?

Kim Davidson:

And we're looking, stay close to your customer, right? That's really the trick in everything we do.

Marc Petit:

Yeah. And in terms of being nice to customers, a year ago I think it was, in January 2021, you made the Houdini engine free. So what kind of impact, why did do you make this decision, and what kind of impact did it have for SideFX?

Kim Davidson:

Well, I don't think a lot of people understood what Houdini or Houdini engine in particular could do. We had these few licenses at every studio, but they weren't even being used that much.

Kim Davidson:

And they weren't expensive, but by making them free, we said, okay.... And we did that the start of the year last year. And we were saying, okay, let's have a goal of increasing it to two to three times. And it ended up increasing the usage and installation over 10 times. So I'd say, to answer your question, it's going very well. And it's been well-received. And more people are creating assets for use in the community.

Kim Davidson:

So I just think that's wonderful. We're all working together. I got the engine, you create. You help me with the tools and I’ll keep working on the engine. I don't know what things people need, because they're so changing and evolving. So having this platform called Houdini and Houdini engine, I think is really the key. So making it free was a big deal.

Patrick Cozzi:

Kim, I admire the strategy. And I think for the metaverse, where all computing becomes so 3D centric, that you need to get it out there and you need that interoperability, as you said. Someone's building an engine, someone's building tools, and everything works together in that ecosystem.

Patrick Cozzi:

So one thing I was very excited about earlier this year was the Matrix Awakens Demo for Unreal Engine 5. It's among the best real-time I've ever seen, probably the best. We were lucky we had Kim [Libreri] from Epic, the other Kim, on the podcast a few episodes ago to tell us about the making of it.

Patrick Cozzi:

And I believe Houdini played a big role there, so I'd love for you to share us how Houdini was used?

Kim Davidson:

Yeah. I played that role. For sure, I'll let you know-

Yeah, it played a role for sure. I'll let Kim Libreri say how big of a role, and my CTO gave me a lot of tips on all the little nitty gritty technical details. But it was a positive experience for sure, working with Kim and his team. We've known Kim, well at Epic Games before that. We worked with him at ILM and Digital Domain. But it was great working with him again on this project. And I'd say they did a lot of the heavy lifting. Our team was doing the Houdini engine updates and support, and we wrote some new tools, the pivot painter tool to help with breaking glass. But largely it was creating that city based on a blueprint and raising that city up about the size of Los Angeles. I know Kim talks about that a lot.

Kim Davidson:

But that's a huge thing that obviously, if you want to kind of re-change anything, you want to do that or build it just in time or whatever, and you want to do that in decorations. And a lot of that's happened. A lot of that mesh placement happens in the Unreal Engine. But Houdini's sort of like, you can test all that stuff out and there's a one-to-one mapping there. And then we just put a point cloud over and then they instance the meshes onto those point clouds, so it's a real handshake. But remember The Matrix Awakens was done in Unreal 5. And Kim talks about, it's a little green, it's new. But they wanted to push all the new features, and by doing that, they're also pushing, changing the engine. And we're sort of telling them what's wrong or what's right or what can work or what we can do with the Houdini engine. So there was a great partnership there back and forth between primarily our development team.

Marc Petit:

Yeah and I think you need proceduralism. I mean, you can't manage by hand a model of the size of that city. So I think it was a great proving ground, as well as we expect to have large scale open worlds in the metaverse that we're going to have to learn how to work at that scale where hand placement of things is not an option anymore. So it was, indeed I observed that. It was a great collaboration. I think it was a lot of learnings on both sides and hopefully people get this dataset at some point, it will be released by Epic, Epic made that commitment. It will be interesting for people to learn from that asset. And I think it shows the way of how we're going to have to manage content at scale.

Kim Davidson:

Yeah, partnerships.

Patrick Cozzi:

Kim, I also wanted to ask you about another project, Project Titan. Could you the tell us the backstory on this

Kim Davidson:

Sure, yeah. So what Titan is is an in-house deck demo designed to production test procedural workflows like we did on matrix while creating a 3D environment for Unreal Engine 5. So it's like the shoes on the other foot, right? We're now the people creating it, albeit much smaller, and we're getting support and we're giving feedback to the Unreal 5 people over at Epic. So it's one thing to support them. It's another to do your own, kind of like live it yourself. So we became the customer. It brought us closer to what our customers are wanting to experience in real-time by creating, but also by creating that IP, we created dozens and of tools that we put out on the web, procedural assets. You can go to the website, SideFX Titan, and we want to share that with community and we want to keep sharing that.

Kim Davidson:

And we did work with a few in house, but we worked a lot with outside artists, because that's how a lot of things are done these days. So we had artists that had never touched Houdini. We built a tool for them and put them in the Unreal environment or put them in our environment or both depending on what they were doing. So that was a lot of good support and feedback to then build the tools, but put them out. So we're going to continue with that. It's a live environment. It's an Unreal environment. We're already seeing some of the people build, use those tools to build their own environments, and we'll call them worlds. That's what the metaverse needs, worlds. But everyone needs to start somewhere. So Titan to independence to building on bigger teams and contributing to bigger and bigger teams, that's what's going to build this world.

Marc Petit:

Yeah, interesting segue because that's what I think is fascinating about Houdini and some of the work that you're doing with Unreal Engine is kind of, it's a window into the future. Because we would want, the metaverse is actually a lot about those large scale, virtual worlds, sometime, maybe most of the time realistic. So you talked about proceduralism and smart assets already. What are going to be the key building blocks and the key technologies to build those? So naturally anybody can build those virtual worlds and how do you see Houdini play a role in that landscape moving forward?

Kim Davidson:

Yeah, well, it's a continuum, Marc. I mean all the way from letting kids have some Lego to having some more sophisticated content creators sort of build those tools or use those tools in a really, really sophisticated and optimize those tools that aren't the developers of the tools, but the first users of the tools or the underlying users. So it's a continuum, I think, to build the worlds. It's going to take everyone, not just the people that are playing and sculpting the world and saying what they want, but also those that say, "This building cannot be moved. This is sacrosanct because I actually sold an NFT." So it's a big thing.

Kim Davidson:

But I think Houdini, really we're behind the covers. We're largely helping those early content creators for these metaverse or metaverses, whatever we want, world or worlds. Creating large variety of assets that populate these worlds whether it's city, a forest to deserts, futuristic, fanciful, whatever you want to imagine, Houdini can help you create that huge metaverse. That's what's needed because the best metaverse is the one's that are going to be most immersive are going to be the ones with the richest variety. So if you can build that, Patrick, you're going to have more people live in your verse than Marc's metaverse or vice versa. Maybe that should be a little challenge for your small little teams.

Kim Davidson:

But I have to say we support those world creators and Houdini in turn. But whether it's Houdini or whatever, we want to stay close to them and see where that metaverse is going. And we don't just want to populate it with fences and trees. I mean, we have a lot of energy behind Houdini effects now, which is quickly advancing Houdini's character capabilities from motion capture, motion retargeting, all the crowd tools which got a technical Emmy award, some physical correct motion based on weight and what a human body might do. So I did want to advance all that, because it needs creatures to really come alive. You need people inside those worlds, too. But we're just supplying the foundation, the tools and advancing those tools, and guys that I think is what's needed to create it. So we're very, very supportive of those world creators.

Kim Davidson:

And again, the more you can do, I think in the metaverse the other thing Houdini might be able to do, I'm not sure. When you play a game, you go from region to region, that data gets loaded and unloaded and then, but it's still there. There's a piece of the disk that's keeping that data and it's always there on the disk. But if you could just have the rules, which are so small and use the compute power to recompute and build that data on just an as-needed basis or a just in time basis, that's where that procedural can play. Like all of a sudden, people are going to a part of world that hasn't been explored for a long time. So it's like, "Okay, build it up. It's not sitting anywhere except for the rules to build that road, that part of the world." So I think that's where proceduralism and Houdini and can play a part. I'm sure the engine providers sort of think that will way as they build the game. They're trying to make the game smaller as well.

Patrick Cozzi:

So Kim, you just touched on a geeky question I've been itching to ask, which is the trend, the state now and then the trend of where the procedural execution is happening. There's edit time and you bake out your geometry and textures. And then there's runtime where you can end up with this really concise rule, but you need more edge compute. So are you seeing a trend that's more towards the latter?

Kim Davidson:

I think so. I really think so because the amount of storage that's needed is far going to outweigh the cost of that compute. And not for everything, maybe. Again, because obviously if I destroyed something in that world and you want that to be persistent, there's no rule for that, really. You'd have to create a machine learning rule to recreate the destruction. So somehow differences to the rules have to be retained, for instance, in certain worlds. So I don't think it's for everything or every world, but I think we're going to see more, like just build it up in real-time or again, if you know, hints where the people are entering that world are going to need that world and you run it in the background, get it all ready and it's available.

Marc Petit:

It's an interesting point that proceduralism, but a big theme about this podcast is about creating an open metaverse and making sure that those worlds, there is interoperability. So how do we think about interoperability in the context of a procedural system? Do you think that we could define standards, share rules? How do we... Because if things happen at runtime, how do we create an open world, a world that is truly open. I'm losing my words. Something that's interoperable?

Kim Davidson:

Right. I think you guys touched on that on a few other podcasts, if I remember, but it's like, "Do you have an engine here and an engine here?" And somehow the right handoff has to be made, so there's probably people like you, Patrick, or others that I've heard that are better than I, to figure out who's going to have the rules. Because I think we want, also, I think there will be several metaverses or worlds, if we can call the whole thing, but because that's how we get competition and advanced tech is if you start locking in too soon, one world can't advance. I like this little space race that’s going on.

Kim Davidson:

But you're right, Marc, if you want to just move in, we don't want an internet where, and I'll go back to the internet example for the listeners. You don't want to sort of have one kind of control C mean something here, and menu is pulled down this way. In the other world, the world's got a whole different UI, different rules, different language. Could look different, but we have to figure out how you can move fairly seamlessly from, in this case, webpage to webpage. But in this case world to world. It's a challenge, but it we all want that, so it'll come.

Marc Petit:

Absolutely, but you guys are strong proponents of the open source and open formats at SideFX. You made a big investment on USD with Houdini. So, how is it to work with an open source format for a commercial software? How is interaction with Pixar, and how are things going on this front?

Kim Davidson:

Yeah, no, that was almost serendipitous in some ways because we have a long term view of Houdini, so we're always updating pieces of the architecture. They're getting a little older, and this is several years ago, we were talking about a rewrite of all our lighting tools within Houdini. We started talking to the Pixar people and they had USD, obviously is the backbone there, but they had a couple other users that were on... It was just early days, so after a bunch of meetings, we jumped on the bandwagon and said, "Let's make USD the heart." It's really ingrained in the lighting tools. And we sort of learned.

Kim Davidson:

And as you imagine, being an early adopter can be painful, but it also has the abilities that you get to shape this thing along the way, whatever it is, and work with your customers to shape it as well, not just Pixar but others who have got on there. Then the bonus, really, is we've got layout and look dev and interface with all the rendering tool things. And we've rewritten our renderer to be USD-centric, so it is karma. That's just a wonderful thing. That alone, just having that renderer really gets everyone into the same ballpark. And the best part of USD, of course, is the U, the universal. Because really, Marc and Patrick, we are in a community that really embraces openness and standards and supports these things.

Kim Davidson:

I saw a lot of like black boxes in that, and we never wanted that. And plus, it's just philosophically, why am I creating, recreating the wheel, a smaller company? It may be expensive, but it's just frustrating. So let's all evolve these standards to the best we can and we'll create it for everyone. And USD is one of these newer standards, but yes, you were asking if it's being integrated. Absolutely. More in the film pipelines than the games, but the games people are starting to figure out what parts of it are very useful for their needs.

Marc Petit:

So what about MaterialX? Is this something that you've been looking at and you think has the potential to also help us with sophisticated materials?

Kim Davidson:

Yeah. We'll have support for MaterialX as well.

Patrick Cozzi:

Kim, we covered a lot of ground today, procedural ground. Is there anything that we didn't cover that you're hoping we would?

Kim Davidson:

Well, the metaverse is like a billion questions, but what we could be talking about, but I should be talking about, I don't think there's any one that I would single out. But I guess maybe I would say, because I don't know if I've listened to all of them, but I do think it's a bit of a hype cycle right now. I'm a big believer, seeing all these hype cycles when you've been around, but that doesn't make it less real, I guess I might say. We've been on this new app before the hype started and we'll be on it when it starts dying down. It is real.

Kim Davidson:

And I think that there are people that would say we're already there. Some systems are already being... social systems are already there, Roblox, that sort of thing. But I think within a decade, just caution everyone that this is early days. But eventually we'll all be in this immersive experience and you'll go, "Wow, so that's what they're talking about in that podcast." Because it's a wow experience to be in there. To be with you guys that's not a video, that is much better. When you've done that, that is much, much better. Yeah, for sure.

Marc Petit:

We agree. And I think we kind of can see that podcast is an antidote to the hype cycle, highlighting once it's been long in the making. You've been at it for 35 years, Kim, and the metaverse has its root in all of that CG technology. And that openness and that interop roadmap is probably a 10-year roadmap, to get to the level of openness that we want. So it's been a long road and it's a long road ahead. But yeah, we get the moment right now. All right. So finally, is there a person, an organization or an institution that you would like to give a shout out to?

Kim Davidson:

I should shout out to the Academy Software Foundation that was formed a few years ago and we are a founding member, but again, they have this neutral platform for open source developers in motion pictures and broader media. I think that's just one example. There's other open sources and standards out there, but definitely we really support and love those guys.

Kim Davidson:

And to my SideFX team past and present, that's my last shout out because they're just wonderfully innovative. They're active in the industry and they're supportive of the customers who are really the creators at the end of the day. And thanks to both of you. This series of podcasts is a great example of you're giving back and advancing and the industry. So, shout it to you guys.

Marc Petit:

Well, thank you. Thank you. I knew you would be good at it because you've done so many Oscar acceptance speeches, so I knew you would be good at that question.

Kim Davidson:

I'll wait for the transcript to go up to see how good. I'll reserve my judgment, but thank you, Marc.

Marc Petit:

No, it's been a pleasure, Patrick. Any last question, any last word for today?

Patrick Cozzi:

Kim, just really appreciate you sharing your knowledge and your perspective and your passion.

Marc Petit:

And very happy to cast a light on Houdini, it's a little bit of a... Not a secret, but you've been behind every blockbuster movie and a lot of those games for the past many, many years. So you are one of those very strong and important contributors to the industry. So, Kim, it was a pleasure to have you with us today.

Kim Davidson:

Thank you.

Marc Petit:

And I want to thank the audience as well. We get continued feedback, mostly positive. People seem to enjoy and our secret is to have good guests because we just put a microphone in front of them and the content takes so naturally. So Kim, you've been one of those amazing guests. Thank you so much. Again, thanks to our audience. Reach out to us on social. Let us know how we're doing, what you want to hear about. Thank you, everybody. Thank you, Patrick. And thank you, Kim.

Patrick Cozzi:

Thanks, Marc. Thanks everybody.