Houdini in Action - 80 Level Podcast

Speaker A: Foreign.

Speaker B: Yeah. So I, my kind of day job is a CG supervisor at an animation studio here in the uk uh based in Leicester and I've kind of been here for about 12 years now at this company and it's Clarend Creative. Um and I kind of joined about a year after kind of graduating from university. I studied um, computer games and visual effects. Kind of graduated, spent about a year working on my portfolio and then um, got hired as a cg as a junior CG supervisor and then kind of slowly worked my way up to CG supervisor and uh, NSC Creative. We kind of focus on uh creating educational science, educational films. So we're actually based at a science uh museum called the National Space center. Um um which is kind of a science museum. Go downstairs and it's full of kids running around from school groups um learning about space science. And then we make shows for the planetarium that's based here. Um but then we also make shows for planetariums around the world. Um, beyond that we work on theme park rides and kind of any kind of large, non standard kind of format films, projection, mapping, Pepper's Ghosts, all these kind of things you see at kind of theme parks and uh, immersive rides. And then more recently I've started to take on some more kind of freelance work as a bit more as a technical artist focusing on Houdini and um, kind of more real time video games. Unreal Engine. Um and then that's how I kind of got started getting involved with um SideFX and eventually uh Project Pegasus which is their kind of latest in house technical educational demo um to kind of showcase how Houdini Engine can be used with um uh Unreal Engine to kind of create a large ah, open world type environment.

Speaker A: I think Houdini has done this tremendous jump from being perceived as a software for VFX and film and uh, become a sort of like a staple in the video uh game production uh pipeline. Um, can you talk a little bit about the things that you think help facilitate that? Why do game developers embrace uh the software and what does it provide that maybe other pieces of software do not? Why is it becoming so popular especially in these cases where you need to build something large very quickly?

Speaker B: Yeah, so um, I think because it is because you can keep everything so non destructive and procedural. Um how I got started with Houdini or just creating personal projects inside Unreal Engine was just the uh, nature of it being very kind of non destructive and procedural meant that working in my spare time on personal projects I could Iterate very quickly, work non destructively. Which meant if I needed to make a change or you know, do something differently, I could very easily make those change and kind of have those changes filter through right into kind of the Unreal projects. And I didn't have to kind of waste time redoing work that I'd already done. Um, and it just kind of kept growing from there, kind of. That's why I kind of latched onto it. And every time I kind of came back to trying to do something in 3DS Max, or whatever 3D software you use, I'd end up having to redo work if I wanted to make a change. And with Houdini I didn't have to do that or more often not. I could um, reuse what I'd already done or even duplicate the same setup, change a few parameters and get a different result. Um, so kind of as an artist level, that's why I kind of kept coming back to it. Um, kind of even just starting out, if I was sculpting an object in Zbrush and I wanted to generate a low poly UV model, I could set up a network that would automate that process for me with a few clicks. I can get that from zbrush, optimized, kind of baked, and then view it inside the Unreal engine in a few minutes. So I really latched onto that and then just started kind of learning and growing things from there.

Speaker A: Could you talk a little bit about the kind, um, of like, what are the big things or themes in Project, uh, Pegasus that you've been working on? What are like, what's it center around? What are the things that are, you know, educational? Uh, what people can kind of. Why would people be enticed to kind of like download it and try playing around with it? Where are like the pieces and building blocks of this project?

Speaker B: So when we uh, started Project Pegasus and when I kind of first was asked to uh, um, kind of start the ball rolling and coming up with what Project Pegasus was, we kind of, we knew we wanted to do an open world style environment because it had Project Titan the previous year, which was kind of like cyberpunk style, uh, city, urban, uh, environment. Um, so we wanted something a bit different from that that kind of stood out but was also kind of larger in scope. Um, so really from the get go we knew we wanted to do an open world environment, um, and probably go a bit more for the kind of medieval fantasy just to have a bit more stood out a bit more from Project Titan that was this kind of cyberpunk environment. And so when we began, or when I began it, I really kind of wanted to focus on creating an environment, coming up with the concept of what the environment would be. Um, and then that would kind of then give us a list of these problems that we kind of needed to solve, kind of to reach that goal of creating this kind of, um, large environment. Um, and I kind of started to draw. Draw up a list of the kind of the areas that we would focus on. So it kind of boiled down to um, the landscape itself kind of scattering things like foliage and stuff. So we wanted to call foliage in some way. Um, and then have um, some kind of village environments. So architecture kind of generating procedural buildings.

Speaker A: Yeah. When you look at it, it's fascinating, right, because it's ah. It's a huge area. Uh. Right. I don't. I think we have some numbers in the article. We'll leave it in the. In the description. But basically it's like this huge area that is completely procedurally, uh, generated. There is like forests and there is like very complex uh, landscape with. There's like. Where there are valleys and then there's mountains and then there's some rocks and everything. And then you have vegetation and then you have these amazing little kind of like villages, uh, with um. Some kind of edifices there, like with houses and with church and so on. Um, it's fascinating to see all of that being done with basically a bunch of the procedural formulas and the elements. And um. I think the biggest question that people have when they look at it, uh, because they're. They're thinking, okay, um, I'm speaking from personal experience as well. So a. I will never be able to do it because it's way too complex. And uh. Uh. I'm. I don't know how to code or something or there's some other roadblock. And then, uh. It also will gonna be. It's gonna require me a lot of time. So instead I'm just gonna go either try to build it all myself or with a team or I'm gonna go and buy some of the. You know, go to Kit Bash or maybe some other spot and uh. Just buy a bunch of uh, stuff over there and just place it by hand. M. I'm wondering if you could. Because it's a different mindset. You know, it's a different way to approach problem, uh, solving. Um, can you talk a little bit about uh. What's. What's good about Houdini in general and this uh, kind of procedural approach? You talked about that it's non destructive. I think a lot of people don't really understand, understand what it means like in terms of, I think when people are actually builders when they do this uh, on scale and this is like their day job, I think they understand it. I, I talk with a lot of guys from you know, all over the, all over the world. But I think one of the biggest examples I remember vividly is like the Ubisoft where they were working on the Ghost Recon games or Far Cry games. They are doing so much work there, it's just, it is just impossible to try to cram all of that even in the, in the team of their size. They have like huge teams, huge company offices all over the world and still they have these three uh, you know, Houdini guys who are like incredibly smart and like they, they spend some wood time with, with the software and then they create this amazing, amazing map already populated. And then the game designers, everybody comes and start kind of playing around with it. So talk us a little bit about like this almost like a philosophy behind uh, this approach where you use the software to create this kind of like you know, scale very quickly and then start adding more stuff to it or you know, art directing it and so on. I think this is something that's missing uh, in some people's minds. Perception.

Speaker B: Yeah. So I mean we definitely took a lot of inspiration from like the Ubisoft games like Far Cry or Ghost Recon and I kind of played a bunch of them when I was starting out just to kind of get, get some ideas of what kind of things we would kind of um, cover, um, the project itself kind of grew quite organically. Um, we didn't have like a set path that we were going to follow. Um, so in terms of kind of trying to kind of give an overview of the way the kind of the environment's built. M so kind of using Houdini was kind of able to basically block out this island, this 4 kilometer island kind of. I mean Houdini needs a big kind of toy box of different tools that you can kind of just put together in any different order and kind of build whatever you want. Um, so we can kind of create basic kind of block out shapes to generate the kind of shape of the island and then use kind of procedural erosion tools to kind of stimulate um, rainfall and precipitation which erode the landscape and create rivers and valleys. And then as you're going, you're producing kind of extra information and data from, you know, where it's calculating kind of waterfall in these rivers. You can use that to drive placement of um, trees or rocks or cliffs. So we can kind of use the kind of erosion and the debris kind of fields that it developed, that it generates to place rocks um, in these areas or um, kind of spawn um, particular types of trees kind of higher on slopes and then different kind of trees lower down in the valleys. And as you're going, you're kind of building these kind of rules almost that govern the placements of different objects. And uh, because it's all non destructive you can kind of go back to the start, make a change, change the shape of the island and kind of have these kind of filter through um, and kind of update the placement of objects. George, uh, Hume was the artist who focused on the landscape. I kind of blocked out the initial of block out and shape that I wanted and then he kind of took on the, took on the landscape and built it up into the kind of very detailed landscape that you see in Project Pegasus and scattering all the objects and um, trees and paths and roads. So you kind of start to build these kind of layers and layers and layers of different tools.

Speaker A: I am fascinated by it. Like I, I, it's uh, because I've seen it being developed for like, for years and it's always nice to see that with every release there's some kind of a new feature. And I remember when they, they introduced this erosion element and you, you build the rivers and everything and everything's done like by magic. You basically click a button, then it just like starts working on the screen. That looks fascinating. And uh, I love the, the floor tool. That's also exciting uh, to see how this vegetation is all being done within one software because that's, that's kind of insane because if you, if you go to you know, any school, any school that does cg, uh, kind of like whatever, or if you even go to a large company like something like in here in California, you would see people who are saying like I'm a vegetation artist and their whole thing is that they're, they're doing vegetation and they are almost applying this kind of like a Disney approach, uh, where you know, round is good, triangle, uh, is bad and you know, sharpie is bad. And they kind of build the vegetation like very um, you know, by hand, super attentively, uh, kind of like trying to make sure that it's all kind of like done neatly and so on. But now you have these tools where you can just try to learn it, figure it out and then massively Produce at scale, like building these incredible worlds. And when I think about it, like I'm trying to understand what does it bring to game development in general. Like if you look at the stuff that you've done, it's basically a game. You basically created a game. Like if you add some kind of gameplay, logic and other elements like monetization or whatever, you have like your own island in Fortnite or something. Right?

Speaker B: Um, yeah.

Speaker A: I'm wondering how do you see these tools, um, Houdini and maybe some others as well. How do you think they're going to influence uh, game development in general? What are we going to see? Do you think uh, people will start embracing this technology even more? Um, not just large companies, but maybe smaller companies as well because makes so much sense for a smaller uh, studio, you know what I mean? Like I think Ubisoft for them it's uh, just Ubisoft that any kind of like all these large, they're like, okay, we have the resources so we can invest, we buy these license and we get this guy who's gonna tinker with it and maybe it's gonna, you know, bring some fruits later on. But with smaller studios they're like, you know, a little bit on the fence. Like maybe they're reluctant to do the investment, maybe they uh, don't have the time. But um, just like from the high level perspective it seems like it's a perfect tool for them because you, you spend some time with it, you learn how to do it and then you can create content very quickly. And this is like the biggest cost um, factor for, for a lot of those studios. What do you think, um, you know, how do you think it's going to develop? Do you think we'll see, you know, embracement of Houdini in like uh, you know, maybe smaller or mid tier companies, um, or maybe maybe in other industries. Like what's, what's your take?

Speaker B: I mean I think we already do see lots of um. Well I've certainly seen Indio studios using or leveraging things like Houdini because you can. Much like I was just saying about working on personal projects, you can work um, in a procedural way and generate variations. You know, create a, um, set up a network like it is a recipe and then change a few parameters and you can generate kind of lots of variations. Whether you're generating, I don't know, rocks or landscapes or uh, whatever it is, you can kind of very quickly create different variations. And so kind of, yeah, for indie studios I think being able to leverage that kind of workflow means that you can have a smaller team produce more content. Because I think regardless of the size of the studio, very often the. The expectation from the audience is the same, right? They still want an open world to often explore and do all the kind of fun things they usually expect to do. Um, so the tools like Houdini that can um, help generate content in kind of a more streamlined way, um, means you can kind of meet these expectations. Um, and I think, I mean we as a, our team, we were kind of only four of us, right, trying to create this very large environment. And hello, I was on the project for about eight months. Um, but really we kind of maybe had three or four months just focusing on creating the environment itself. And the rest of it was kind of more on um, the kind of documentation and turning it into learning content. So we had to create a large environment in a very short space of time.

Speaker A: Um,

Speaker B: but by using Houdini and you know, I could essentially create uh, a tool that can automate a process. Um, like you said, the um, Flora tools was you know, built with a desire to create more realistic. Create trees a more realistic and natural way that kind of maybe the current um, kind of Houdini Start Effects Labs tools, um, kind of fall down a little bit on was to kind of kind of build and maybe prototype a new way of modeling trees. And that was a tool that the guys kind of, that um, Fiker and George kind of prototyped over the course of um, a month or so. So kind of there on a small scale as well. We're prototyping tools uh, that an artist could then take and reuse to create the work that they need to. So again that's kind of like allows a small team to kind of create these very useful tools that artists want and need. Um, and then they can take them and kind of run with them and um, create the work that they need to more easily.

Speaker A: Ian, to kind of like um, wrap this up, um, and talk about. Because I. What I like about these conversations um, with like technical specialists like yourself, uh, as well as doing these things at 80 level is that uh, you kind of take this highly theoretical kind of like uh, math heavy um stuff and you turn it into something tangible and um, like adding like a little bit of a wow factor. Right when you first see like an algorithm generating uh, a building and kind of giving you like variation after variation after variation, you can't escape kind of thinking that this is some kind of like witchcraft or something. Um, and uh, my question is like when you look at this project. Um, what are the bits for you and maybe for your team that worked on it? What are the bits that felt like witchcraft where you worked on some of them and you, you know, you put you, you execute and then suddenly it's like, wow, I didn't expect this. What were like those kind of like little discoveries that you made along the way?

Speaker B: I think, I mean, uh, because when I started the project with just myself kind of blocking out this very basic environment with cubes and basic shapes. So I was very fortunate to kind of end up with a team that I had that kind of was able to take the bits that I was kind of um, messing around with, trying to go with ideas and kind of take it and run with it. And um, kind of put the hard work they did to kind of create the inform that it was. But um, I think it's, it's always when you look at something procedurally because you are, you know, you're not designing every kind of square foot by hand. You are generating it. So there was points where towards the end of the project I kind of spawned the environment in Unreal to run around and check out uh, what it looked like and just make sure it was working. But I'd end up just kind of following a procedural path, not knowing where it's going to end up and um, kind of eventually just kind of following it to the top of the mountain and kind of looking out over the landscape or kind of running down and finding kind of a valley that was kind of full of, of trees and, and a rock slide. And nobody had kind of stopped to think about designing these kind of little unique areas. They just kind of spawned out of the, the rules that um, were, were set in place. So I think it's those kind of things that as you know, you're not expecting them, but they've kind of just appeared in the game as if by, as if by magic. Um, and they suddenly feel like, kind of like real places. Um. And uh, there's definitely a point like when we work on the project and you know, it's progressing, but um, there's always kind of an ugly point in any kind of creative project where you're ah, not quite there yet. It kind of needs a few more iterations and things just haven't quite clicked into place. And then when they do and kind of, you've got the kind of right level of detail and the right um, kind of level of polish and all of a sudden it just kind of all kind of comes together. Um, and yeah, we kind of had this world that we could just kind of start wandering around and exploring and suddenly kind of felt like a bit of a living, breathing place.

Speaker A: Yeah, it's almost like you are discovering what the. What the. What the machine has done. Right. It's sort of like. It's almost kind of like giving you these building blocks. And that's what's um. That's what's amazing. Like, I think I heard with uh, I heard the Guerrilla Team, when we're doing the first Horizon, that they had the same thing where they um, it generated so much stuff the. For them that they were discovering the game for themselves like that they were not really going in our directing every little bit in piece. And I think this is some kind of, um, this is really a unique element to what this kind of software and this kind of pipeline, um, brings. And I, uh, think it brings so much more surprise and uh, kind of like, um, a little bit of this creative uncertainty that is revitalizes this whole process. Because so often we are. Especially if you are like in ufo, VFX and things film. I mean you're working on this shot for what, like months, sometimes years, right? And then it's almost like you're. You don't really see this thing and then you see it for like a couple of seconds when the film arises. And here with like clever tools, you can uh, create this discovery almost every, every single time and try to iterate and so on. Um, Ian, thank you so much. I think this is like an amazing project and just as like the previous urban one that Houdini did. It was like a lot of fun to explore. I think this one is even more, um, kind of big on scale and so fun to see and understand how all those little bits and pieces work. Uh, so thank you so much for that work. We'll leave the links in the description. People can review. Have a look. We also did an article just in case we can put it there. Uh, good luck and thank you for this, uh, conversation.

Speaker B: Thank you very much. Thanks for enjoying another episode of the

Speaker A: 80 Level Roundtable podcast.

Speaker B: Check out upcoming episodes on the 80 Level website at 80 LV. Join our career site at 80 LV RFP and share our podcast with friends and on your social networks.