Art director Grant Freckelton and director David Peers talk about the studio’s groundbreaking open-source USD ALab, its expansion into ALab Phase 2, and the Unreal Engine-created short film that grew out of their R&D efforts.
From Peter Rabbit to the LEGO movies to the Marvel Cinematic Universe, animation and visual effects company Animal Logic has established itself as one of the world’s leading independent creative digital studios. In August 2021, the company started a new chapter with the open source release of USD ALab, the first real-world implementation of a fully realized USD production scene, from global assets through to shot outputs, including referencing, point instancing, assemblies, technical variants, and shot-based overrides.
While the initial motivation was to create something the studio could use for demonstration and presentation purposes, with the open-source release, ALab became a resource for the worldwide animation community.
Meanwhile, Animal Logic continued to expand and enhance the toolset, which culminated with the August 2022 release of ALab Phase 2. The new iteration includes over 300 production-quality assets and two animated characters, Goggles and Hinge, with looping animation and baked procedural fur and fabric.
But wait, there’s more. To show off the suite’s complexity and quality, and introduce the new characters, art director Grant Freckelton directed a lit and rendered sequence, which everyone liked so much that the team decided to make a very short film based on it. The very short film gradually morphed into a somewhat longer short film, and soon was reconceived as a for-real short film. Directed by David Peers and created in Unreal Engine, Unhinged is due to be released … soon. It’s also worth mentioning that the Animal Logic ALab data set is one of the first digital assets adopted by the Academy
Software Foundation’s Digital Production Example Library and is now available to access via animallogic.com/alab.
In two separate interviews, we spoke with Freckelton and Peers about the ever-evolving ALab saga and what the future holds for the initiative, now that the studio has tasted the fruits of an “Unhinged” short film and real-time production project… with a sock puppet scientist.
But first, giggle through the trailer for Unhinged:
AWN: From the release of ALab Phase 2, to the creation of an extended animated sequence and the current production of a full short film, there’s a lot going on here. Just to provide some context, how did this all come about?
Grant Freckelton: A few years ago, we realized that we were doing a lot of testing in a very ad hoc little setup that inexplicably had snow and a little gingerbread house and a weird 3D version of the Animal Logic logo. And that was how we were putting ourselves forward in any sort of presentation. If we were testing something, there would just be this silly Animal Logic logo in a snowy environment. And we came to the realization that we needed something that was a little more cohesive and was more representative of Animal Logic. That's how the ALab project started.
Felicity Coonan, who's another one of our art directors, reminded us of a thing called the Animal Logic chemistry set, which was basically packaging for our show reel at a certain point. That planted the seed of a laboratory, which made sense in so far as this is an R&D tinkering space. Felicity pitched the idea of a precocious lab rat, which evolved into the ferret character of Dr. Goggles. But when we pitched the idea for the ferret to production, they said, "There's no way we can pull this off with the time and resources that we have. We're going to need to find ways to simplify it."
And so, as we were trying to come up with solutions that didn't compromise the fun of the character, ultimately we came up with the idea of just rendering it as a puppet. And that allowed us a couple of things. One, I thought it was going to be more fun to have a derpy puppet than it would be to come up with a fully rigged character. And it also created a challenge for rigging. I know that we can pull off photorealistic animals, and I wanted to challenge us to think outside the box and come up with something that is deliberately simplified and deliberately rigged in a way that isn't just a straightforward quadruped.
We decided to just give her simple button eyes that had no rigging on them whatsoever. I specifically said, “Do not make them blink. Do not put brow controllers above them.” Because when I'm discussing stylization with directors, they're always like, "Oh, I want the eyes to be so expressive." Which is fair enough. Eyes are the windows to the soul. But then I show them Kermit the Frog and say, "This character is pretty expressive. He's one of the most beloved characters of all times, but his eyes are ping pong balls.”
AWN: So, in an effort to upgrade how you present your R&D work at conferences, you created these characters and this laboratory environment. And then you sent it out into the world.
GF: Right. We use that lab environment internally to test various tools. We have a daily viewer where we can see what's being checked, and we can see the assets that we're creating. But what's exciting is we're also seeing those assets being utilized externally, and people experimenting. I'm sitting on the Blender site watching a thread - of ways Blender can import this environment better - in one of their bug reporting threads, and it's happening everywhere.
AWN: And are people able to really dig down and see how this stuff is made?
GF: Yes, particularly in terms of understanding how USD is set up and utilized within the company. There are all sorts of insights into how directory structures are set up and how each file relates to other files. So, it's not purely a fun, creative thing. There's a technical aspect to it as well. It works on both levels, and it works for big companies and also for the home user.
Check out the ALab promo:
AWN: What’s the advantage for you in releasing ALab as open source? This is something that you've spent a tremendous amount of time, money, and effort to develop. What do you get out of it?
GF: The value to anybody who releases anything open source is the value that comes from people utilizing and learning from it. If the release results in 100 really smart people learning more about the way we do things, then those smart people are going to develop their own tools. Or they may be better qualified when they apply for a job at companies that utilize USD. It's altruistic, but it's also advantageous for everybody. We will benefit from any technical development that comes from what we've released.
Pixar released USD to the world for the same reason. Everybody's going to benefit from a certain amount of standardization.
AWN: More directly, how does this project integrate with and benefit Animal Logic's commercial ventures moving forward?
GF: In the process of designing this environment back in 2020 when we started it, I picked up techniques that I'm now using like crazy on the film that I'm currently designing. And again, when we released Phase 1, people who wanted to use it had to iron out some bugs in order to import the USD to Blender. And now, when I use Blender in the art department and the matte painting department, I can import assets internally in a piece of software that otherwise would've required a show contract and an R&D ticket and internal resources, and so on. So releasing Phase 1 directly led to benefits that have helped speed us up and allow us to do things more easily.
AWN: In our interview, Grant Freckelton told us about the history of ALab, and talked about its release as open-source and its use in Animal Logic R&D and public presentations. How did it morph into its latest incarnation as a short film?
David Peers: It was kind of nonlinear, based on the evolution of a series of things. First, as ALab evolved, it started to take on a life of its own. As I’m sure Grant told you, the environment acquired so much richness over time – it's just full of all these beautiful little visual gags and things. Along with that, Animal Logic is making quite a lot of animated movies now. We have a pipeline that's really great for making animated features, but there'd also been discussions about real-time and things like that. Part of the idea of ALab is it can be used for a lot of different things, so we decided to put it into Unreal and see how that worked.
Initially the idea was just to do something simple. Then I came up with the story for Unhinged, and we released the trailer for it at SIGGRAPH. And we kind of fell in love with the characters, and the story developed more, and basically I just kept working on it. Then we had a gap in one of our feature projects, and so we put a crew together to work on this thing. So we were working out how to make a film in Unreal, and working out the story at the same time. All those things happened together. It's certainly been challenging, but it was also exciting, and people really stepped up to it.
AWN: What was the actual process like bringing Unreal Engine and ALab together?
DP: I should say, first of all, that Epic was very supportive through the whole thing. We checked in with them fairly regularly, and they had people on hand when we ran into problems. We started out with what we called an “Engine first” approach. The primary thing for us in doing the project in Unreal wasn't the ALab part, but seeing how it affected the filmmaking process. One of the big wins was multi-user editing, where we could have a room full of people, or a virtual room full of people, all working together to put a sequence together. We could scout the environment together, we could set dress together, we could make decisions on the fly and act on those in a way that just doesn't happen in a traditional pipeline. So, on that front, it’s fantastic. It gives us the best of both worlds – immediacy and collaboration, as well as the ability to then go in and refine the little details.
We initially tried to do everything in Engine – build, animate, cut, you name it. We were working on the latest and greatest Unreal 5 the moment it came out, we were using the very first versions of their control rigs for the characters, and we were doing most of this from a standing start. Also, we were doing it all in the cloud, so it worked with the whole Amazon Nimble setup. A lot of companies that do these things in Unreal spend a few years building the scaffolding and gradually do more and more over time. They might start just using it as a renderer, and then they’ll do a bit of this and a bit of that. We just went the whole hog from the beginning. And only a few of our crew were that Unreal savvy when we started.
So we of bit off a lot and had to chew like hell, but we got there. For some things, though, we found we just had to go to a more traditional process in order to have something ready for our SIGGRAPH presentation. With animation, even though we built full control rigs and we could do everything in Unreal, the animators, for a lot of the shots, were working on their traditional rigs and then were pulling it back into the Engine. We just had to make some tradeoffs at different points, but we got to see how this affects our process, and where Unreal might fit in the future for things that we do.
AWN: It seems like the minimalist approach you took, which was dictated at least in part by the need to get something done quickly, led to some great design decisions. Will you continue with that aesthetic for the whole film?
DP: Yes. The idea has been to lean into those limitations, because it makes for quite challenging storytelling, in that we don't have dialogue or facial expressions to lean on. It all has to come through cinematic language, physical performance, sound design, and so on. And that minimalism is all part of the appeal. It's like Hinge, as a robot, is more inspired by 80’s animatronics that are a bit janky and don't have all the capabilities of the fully CG-articulated things that you see in contemporary films. And Goggles is very much coming from a Muppets-type approach. We really agonized over the animation to make sure that the character didn't do stuff that you couldn't do with your hand inside a sock puppet.
There's a real appeal to that, and it's also particular to the kind of work Animal Logic does. There are the California companies – the Disneys and the Pixars – who do that big dewey thing, which comes from a classic cartoon animation tradition, and they do it exceptionally well. We have more of an Australian sensibility, which comes out of independent filmmaking, rather than from cartoons. So it's much more an analog, tangible, handmade feel that runs through everything.
AWN: When can we expect to see the completed film?
DP: The short answer is I'm not entirely sure. When it started out, it was going to be 90 seconds, then it grew to three minutes. It's now getting up towards five minutes. It's partly because, as we find these characters and we're fleshing them out and looking for more opportunities for comedy, the film itself is developing. And while Animal has made a lot of feature projects, this is the start of what will hopefully be a lot of other short-form projects, and we want to do it right.
Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.