Check out www.VFXRates.com
Episode 91 — Round Table with Image Engine’s Creature Supervisors
This is Allan McKay. Welcome to Episode 91! I’m speaking with Image Engine’s creature department. This is going to be an awesome Episode!
Image Engine is a Vancouver-based film studio that’s been absolutely crushing it: from Jurassic World to Game of Thrones, to X-Files, to Elysium, Chappie and so much more! They’ve got a massive history as one of the first VFX studios in Vancouver and for being around for a long time, longer than before the boom of 2011. Fast forward to 2017, they’re still killing in as leaders for doing realistic creatures and high-end visual effects.
It’s been cool to sit down with some of the Supervisors, from the Creature Department to Animation Pipeline, Effects and a lot more. This is a lot of fun! I’ve gained a lot of insight and gained some understanding about what sets them apart. Some of the core challenges they run into on every feature has allowed them to grow and expand.
Let’s dive in!
ROUND TABLE WITH IMAGE ENGINE’S CREATURE SUPERVISORS
Image Engine is a high-end visual effects studio based in Vancouver, BC, that specializes in creature design and animation for feature films and television. Founded in 1995, the studio has worked on films like District 9 and Chappie, Independence Day: Resurgence and Jurassic World; shows like X-Files and Game of Thrones (Seasons 5 and 6); and much more. In 2010, the company was nominated for an Academy Award for Best Visual Effects, for their work on District 9.
In this Episode, Allan McKay moderates a round table with the Creature Supervisors at Image Engine. They discuss the company’s history, work flow and pipeline; and their commitment to furthering the art of visual storytelling.
Image Engine’s Website: http://image-engine.com
Image Engine on Vimeo: https://vimeo.com/imageengine
Image Engine on Facebook: https://www.facebook.com/ImageEngine/
Image Engine’s Twitter Handle: @ImageEngine.
[-56:47] Allan: Okay, guys! Thanks again for taking the time out to do this Episode! Just to start out, I’m speaking with:
- Jenn Taylor who is the Animation Lead;
- Barry Poon, the Asset Supervisor;
- Jason Snyman, the Animation Supervisor;
- Andrew Kaufman, the R&D Lead;
- and Shawn Walsh, the General Manager and the VFX Executive Producer.
Just to start out, can you quickly introduce yourselves?
Shawn: I’m Shawn Walsh. I’m the General Manager and the Executive Producer here. I started from the lowly depth of the old CG television animation realm and became a production person; and they’ve asked me to run a show here.
[-56:02] Allan: That’s awesome! Thanks, Shawn.
Jenn: Hello, I’m Jenn. I’m an Animation Lead here. I also started doing cartoons and television, and have worked my way to features. I’ve been with Image Engine since 2012.
[-55:47] Allan: Wow, that’s cool! Thanks, Jenn!
Jason: I’m Jason, I’m the Animation Supervisor. I’ve been doing this for about 15-17 years. My first job was as a matchmover and I’ve worked my way up. I’ve been with Image Engine for about a year and a half now.
[-55:27] Allan: Sweet!
Barry: I’m Barry Poon. I’m the Asset Supervisor at Image Engine. I’ve been working in the industry for a while, from tv to games to film. I’ve been at Image Engine for almost 6 years now.
Jenn: Longer than that!
Barry: 2011? Six years.
[-54:59] Allan: Time flies when you’re having fun!
Andrew: I’m Andrew Kaufman. I’m the R&D Lead here. I’ve been here since 2008, always as an R&D.
[-54:24] Allan: Like I said, I really appreciate your guy’s taking the time to do this. We’ll just dive into the background of Image Engine. I figure it would be great to pick Shawn’s brain about how [the company] has grown over time and how you got started.
Shawn: Sure! The company got started in about 1995. It’s was spun out of a graphics department at a company that delivered video post services in Vancouver. The guys who formed Image Engine originally, were three people who worked at that company. They tried to take advantage of the local growing industry in Vancouver, which was a lot of television work: Poltergeist, Star Gate (the original, longest running series shot primarily in Vancouver).
Those original three grew to a company of 20 or 25 when I joined the company in 2005. It was shortly after that that we decided that we wanted to grow the company from mostly visual effects, HD company into something that could do more than that. We struck out and tried to establish ourselves as a film company. It was a huge transition. Over the following couple of years, we got a bit of film work going, working on things like The Incredible Hulk (the Ed Norton version) and got a bit of notoriety for being a boutique that was doing good work.
And then we had this lucky happenstance to connect with Neil Blomkamp, who was getting going in his career; landed District 9 and went on to successfully deliver that project — which [served] as a relaunch of the company, I think. It’s become our new DNA. Since then, we’ve gone to work on feature films. We still do some commercials. We’ve done television work, with big budgets. This year, we are working on three different tv projects; but the fast majority of the company is built on film work.
Along the way, we’ve built a reputation for primarily being an animation house, like creature animation, with a lot of R&D and development, and supporting the team with the ever-evolving production glue: custom code. We aren’t writing all the code, but between the pieces of software. [It’s] more related to workflow and how we do the work, as opposed to the major pieces. So the cultural aspect of the company are both things: The vibe of the company is built on performances of these characters and creatures and the tools of technology; [as well as] creating an environment where people can actually do the work. We try to make it so that the artists actually have tools that they can be their best work with.
[-49:50] Allan: I thinks that’s really important! I always look at the 3D packages out there as being generally aimed at everyone, not laser focused. The flip side of working for major studios with proprietary software is that there is major upkeep and it’s hard to keep it going. As opposed to having a team to build a bridge between what the software is built to do and what you need to do. You’re building additional pieces that are built specifically for a project, and you’ll be able to tackle whatever that project needs.
Shawn: Recently, we didn’t realize that in developing some of our own key pieces of software, that it would have a benefit in our ability to go back to television because the turnaround time was something that we feared. Andrew — who is in the room here — gave a talk to our teams. One of the comments we took away was that there had been a transfer of technology to projects like Game of Thrones that enabled us to behave like a boutique again. So some of the technology has trickled down in a way that we didn’t anticipate when we started building a sophisticated film pipeline.
[-47:52] Allan: That’s awesome! One thing I want to touch on quickly. I used to live in Vancouver in 2009. I was coming from film and was a bit depressed because there was hardly any film work going on, like you. As soon as I left, the place blew up into one of the biggest VFX cities in the world. You, guys, have really been dominating the creature area. What has your experience been over that time, for the last 10 years, as Vancouver has changed? How did you set yourselves apart?
Shawn: It’s been a really interesting journey. That release year of 2009 had:
– The first Iron Man that the Embassy VFX contributed significantly to.
- Watchman that MPC was doing when they had the first converted technicolor corrective services.
– And we, of course, had worked on District 9.
That year was a bit of a watershed year. This super high-end work was done for features not just for commercials. But since then, we grew from 25 to about 250 people. We found out that being an independent company, that was it for us. We didn’t really want to continue and push the envelope for growth for growth’s sake. Meanwhile, a lot of the larger companies came into town. It’s been interesting. We’ve only seen the benefit of that: There is certainly a lot of competition, but the artists’ pool has grown so much in the city that we’ve maintained this nice mix of people who have been pretty committed to us and been here for significant periods of time — and then having about 1/3 of the staff who are new to us and bringing fresh ideas and mixing with our folks, and we love their commitment. Looking back, there certainly has been a huge amount of growth!
[-44:56] Allan: I completely agree! It’s not really about the competition but about the growth that has happened for the past 8 years. It has put Vancouver on the map as a giant hub for film work. Even before that, if you’re in LA, you could get work. If you weren’t, there was a bit of a resistance when you’re trying to convince studios [to hire you]. Now, there’s definitely been a requirement to have a presence in Vancouver. You, guys, have your roots there.
I figured it would be really fun to nerd out about the process and the pipeline. You, guys, have such a strong creature focus, as well as having a 360 production presence with FX and compositing, digital environments. It’s definitely all the tools that give you the strength as well. I was talking to Mark Theriault about all the visual effects you were doing for Chappie, the procedural chains sorting out for all the shots. A lot of what you do is really genius: You have the artists do what they do; and the bits they don’t need to do — you can automate. That way artists get to focus on their strengths. In terms of your pipeline and workflow, what are the typical tools you rely on? Is there any specific technology you find to be the backbone for what you do.
Andrew: Yeah, in terms of all the animation stuff, that’s kind of started during the District 9 days when we were trying to build a film pipeline from scratch. So we started building this project called Cortex, which is an open-source project which was trying to provide the backbone that does the glue work you were talking about. And that has evolved quite a bit over the years. And then, there is this other open-source project called Gaffer, which one of our ex R&D Leads started in his spare time and we were using it for asset management, actually. Those are the big open-source things and there are two or three internal projects that have names, and a lot of glue code and plugins for individual purposes. But a lot of the management is run by Gaffer.
[-41:48] Allan: Cool! And in terms of regular commercial software, do you use Maya for animation?
Andrew: It’s all the big apps you can think of: Maya, Houdini, Nuke, ZBrush, 3DEqualizer, Mari. It’s pretty much all the leaders in that category. We’re using Arnold for rendering now. It hasn’t always been the case. And lots and lots of added value either via plugins or the glue between the apps.
[-41:11] Allan: I was speaking with Chris Harvey the other day. He mentioned I should ask you about your look development pipeline. It’s exceptionally impressive what you guys are doing. Do you want to shed some light on that?
Andrew: I assume he was talking about the Chappie days. That was all done in Gaffer as well. We were using 3 light as the renderer at that time. It’s kind of a procedural workflow for assembling shaders. I think a lot of the cool work on Chappie — which Barry knows more about — has to do with the access we had in terms of pre-production, designing the character upfront, before it was even manufactured for the shoot and getting realistic lighting and matching it.
Barry: Yeah, I would say that’s probably the biggest thing that Chris was talking about: How much time we spend on developing the look of an asset in the lighting and shading side of things before it actually [went] into production. So we spent a lot of time getting it in a neutral set-up so that any properly graded HDRI in a shot that we drop the asset into is just going to fit in properly, give it a really good base for the lighter to work with. We probably spent a little bit more time, especially when we get into the hero assets, making sure that they’re going to hold up; so that once they’re in production and those assets don’t always come back to our department for fixes, but there’s obviously things that do come up that aren’t expected, or we need to tweak something here or there.
On Chappie specifically we did spend a lot of time getting those characters to look great and not very many things came back. There were some things at the end that we had to do different variations on, but the characters held up really well for most of the show. Which is really helps us: We aren’t a huge studio and we only have so many resources to throw into a show. We need things to work efficiently by when they’re needed. We can’t afford to have an asset hold up a show for a week or two, or even a couple of days.
[-38:49] Allan: That’s really cool! On the subject of that, what was the typical production timeline from start to finish on a project like Chappie and how big was the team?
Shawn: Chappie ran for the about 18-20 months if you include everything. I think at maximum, our team was in the 70s: 70-80 people. That was when animation was in full production and shots had finally began: you have assets, shots in animation and you’re starting to light and do the comps for the first set of files. But in terms of the people who worked on the entire project? Probably 140-150 people, something like that.
[-37:54] Allan: That’s amazing to be able to turn out so much work in that amount of time, and keep organized all the way. I’m always curious about this stuff: You, guys, do really insane amount of creature work. When it comes to shooting, it seems pretty seamless to get an actor come in and do their performance, and you can seamlessly replace him with a digital character. What typically goes into the preproduction when you’re going out on set and shooting a character? Is everything match-moved or what’s your process?
Shawn: If you look at something like Chappie, there is some additional specific things we did. In general, it’s a combination of things we do with production:
– What kind of stages are they shooting on?
– What sort of locations?
– What’s the camera style?
– What are the Director’s and the DOP’s intentions in respect to lens choices?
We have to get a basic understanding of what type of a film they’re trying to make from a cinematographic perspective. And that translates into: If you want to shoot things this way, we need to make sure we grab HDRI’s in a certain fashion, light our locations, measurements of sets, calibrated photo booth texture. [There is a] standard list of date acquisition that we work on with the clients, to make sure they’re on the same page. In some cases, it’s pretty straight forward. In some case, there is a lot of back and forth conversations about how we are going to gather all the data we require.
In case of Chappie, there was one step further in terms of the preproduction perspective. When we design the character, we went through an additional step: We received the 2D concept design, had approval, we build the first-pass model; at the same time we cyber-scanned Sharlto [Copley] and then refit the design to his actual proportions. So we were working with digital representations of him that were extremely accurate. We then ran motion tests using caption data, so we could then adjust the model, so that we could translate Sharlto’s behavior more accurately. And that’s the information that Andrew was talking about the on-set fabricators. We then recaptured their work in a calibrated lighting setting so that we could then do what Barry was talking about: the accuracy of matching the digital version of Chappie to the practical version.
That’s the sort of measure of specificity we go to. It comes up every once in a while. But there are also projects where we’re kind of, frankly, winging it. We have to be very inventive because there is no data to capture. So, with something like Fantastic Beasts and Where to Find Them. We really have to think outside of the box and create an environment for which there are no measurements. It really ranges. It’s part of the fun of visual effects: You can put as much effort and intention into processes that are so diverse.
[-33:27] Allan: That’s really cool! Talking about Fantastic Beasts, what were the contributions you did to that show?
Jenn: We did a number of the creatures, the main one probably being the Swooping Evil, and the Graphorns (which were the cat-like creatures with tentacle faces). We had a small Owl kind of a creature and the Dung Beatles. The Dung Beatles were actually great because it gave us a lot of freedom to explore animation and give them our own ideas on how [the creatures] would move, especially with the Graphorns and the Dung Beattles because they aren’t an existing animal. So we got to play around with that. It was fun for the team!
[-32:22] Allan: That’s a good point: You have creature work and then you have creature work freeform with which you have more creative freedom, unlike having to replace someone else’s performance. It must be challenging and fun at the same time.
Jenn: It’s kind of a nice balance! Neil Blomkamp’s films are very organized. He knows exactly what he wants, he has actors on set. It helps us a lot because we’re on task right away and we’re able to push through the work quite rapidly.
[-31:36] Allan: What about The Thing? You did a lot of creature work on that too, right?
Shawn: This was a really under appreciated film at the time. We were really proud of our work. It was very difficult, creative and technical work. Having a creature that’s constantly changing throughout the scene was really unique. In this case, there were many different assets as the creature kept changing throughout the film. I don’t know if we’ve done anything since then that’s been this spine chilling, kind of out there, gross, weird construction.
[-30:36] Allan: That project I’m particularly protective of because I loved the original. When I was watching the film, I wasn’t 100% sure if it was a prequel or a reboot, until the very end of the film. The problem with CG is that you have to create something that seems real [and believable]. The stuff you were doing, you’d think visually people would reject it. I think you guys really nailed it: It felt so alien and different, but it didn’t come off gimmicky.
Shawn: Well, Matthijs [van Heijningen] had a really strong vision for what he wanted to do and was extremely respectful toward the original. I think there was a lot of clear intension of being in the same world. One of the big challenges was: How do you make a film that’s a prequel to something that so many people hold so precious, but it also works with the modern audience? We did a lot of fixes and replacements, attempts to change to animatronics or practical work that was beautiful from a design perspective, but the studio or the director felt that it should stand up to the scrutiny of a modern audience. You’re really trying to overcome that gap. Visual effect studios are faced with that challenge: We’re responsible for creating something that goes beyond what is captured by the camera. Does it stand up to what a modern audience is expecting? The audience changes every year. That’s something that isn’t talked about a lot, but it’s something that changes.
[-27:57] Allan: What do you, guys, think is the really big challenge that comes with doing creature work: pulling off the perfect performance, or renders and shading? What are the big hurdles that you face — and that set you apart when you overcome them?
Barry: I think it’s a combination of everything. Getting the shaders looking great in the scene but also the small details in the scene. Getting a small character to run around — anybody can do that. But it’s the small details: the muscle jiggle, the nervous ticks.
[-26:55] Allan: I know you worked on Independence Day: Resurgence. A lot of people remember the destruction work. You, guys, did a lot of creature work on that: the aliens. Do you want to do talk about you worked on that?
Barry: That was my first show when I joined Image Engine. I wanted to be part of a smaller team. It was exciting for me! I could push for the type of technology I wouldn’t be able to ask for at another company. With the aliens, we wanted to keep it close to what the original was doing, that it was a sort of a human inside a suite with a scary aspect to it. One of the things we developed was a dynamic tentacle rig, so it could flap them around. It got us a long way. The fact that we had a system to do that [was great]! At a bigger company they would’ve just thrown people to do that. We had a relatively small team and we were able to work on most of the aliens, except for the Queen Alien. We did all that in-house.
[-24:34] Allan: I think it’s a great point. Looking at the classics, it’s one of the things to consider. The aliens in the first one look really cool but it’s the movement that throws it off quite a bit. There isn’t much freedom in that. Fast forward to now, you’re able to give them a better performance. Do you feel like there is anything with CG that you can’t do? Or does it depend on how many people and how much time you’re able to throw at something?
Shawn: I think it’s anything that you want to do, which is kind of good and dangerous at the same time. Nowadays, you can do anything. You just need that artistic eye to analyze what’s happening in the world around and how to make what people are expecting. When people go to the cinema, they have expectations, but if you can push that — that’s special. That’s what we try to push for here!
[-22:41] Allan: Do you, guys, do a lot of motion capture? What type of processes do you rely on?
Shawn: We go two Xsens suits in-house which is awesome. You just put the suit on, a couple a markers, run out the kitchen pretending to be an alien.
Jenn: Jumping off the stairs, fooling around.
[-22:11] Allan: So, you just put the suit on to have an excuse to act like a kid and have fun? “We’re working, I swear.”
Shawn: I don’t need an excuse! It’s a great way to get an idea straight into the computer, as opposed to spending a week to block it out. Potentially, some of that stuff you just clean up a little bit and push through the pipeline, to see with the shadows. Or, it gives you ideas for stuff you haven’t even thought of before. You might get a jiggle in the chest, and all of a sudden you see it.
[-21:06] Allan: It’s so cool to have that. As animators, you’d shoot yourself on film. Now, you can actually capture your performance and augment that. It’s a really accurate reference. It’s cool that you have that access. Everyone is probably going to have a different opinion on this but: What’s you most challenging project to date, or one that stands out as being worth it?
Jenn: I loved Jurassic World! It was one of the biggest films we’ve done and it’s done well critically, so it was great to have that kind of a feedback, for animators as well. High energy, fun show!
Andrew: For the software department, it was probably Final Fantasy, just because it was a really short timeline and we never really did the whole feature. But it just worked!
[-18:56] Allan: It looks really impressive. You had to do a double take.
Shawn: It [took us] slightly less than a year. There was a long preamble while we were talking to them. Once we got into it, it was about 6 or 7 months of all the CG work, 544 shots.
[-18:14] Allan: How many minutes of animation was that?
Shawn: It was about 23 minutes of the film. The third act, essentially.
[-17:59] Allan: I was just recently interviewing the id Software [www.allanmckay.com/81] and asking them about DOOM. With Jurassic World, was there any pressure to not screw it up?
Barry: I think for sure there were high expectations, but mostly on ourselves. Because for most of us, the first movie was such a big deal. You forget about all those movies in between. You think of the original. We were all excited to work on it.
Andrew: On this one, we did all of their actors.
[-17:12] Allan: I imagine it was all the creature work and the environments. What were some of the biggest hurdles?
Barry: Well, the interactive sequence of running through the jungle was a big challenge. It’s a full CG sequence, with CG plants and they’re ripping through them really quickly. I think that was an awesome we were all proud of, but it was a challenging one.
Shawn: Part of the secret of that sequence was just that we didn’t know the full extent of the CG work. So turning the jets on mid-stream was challenging.
Barry: At the end of the day, we have this really great pipeline to deal with the challenges, but it’s really the artists that help us get through it and the team in the R&D. The experience we’ve had, every show has a different challenge. But without all the great people we have here, there is no way we could pull it off!
Jenn: I think the team that came on was so passionate. That really helped us get the performance we wanted. The people wanted to get it as good as possible. The energy of that kept us going.
Barry: People here are definitely really passionate about their work. One of the hard things is to pry assets away from each department: “That is enough!”
[-15:22] Allan: You know it’s a good thing when artists are un-final-ing shots. That’s so cool! I was really impressed. Did you always approach that scene thinking it would be an entirely CG jungle?
Shawn: Original breakdown had some calculated guesswork. The sequence worked really well as a wide action piece. The studio was really positive about it, we got licenses at that point to make the shots much broader, loosen the camera a bit; show more of the landscape. Be more with the rafters as they were running, instead of observing them.
[-13:58] Allan: You have so many iconic feature films! You worked on X-Files. What work were you doing on that?
Shawn: A mixed bag of stuff. We did some spaceship shots, some set extension work, some compositing. But we also were able to touch the aliens’ grading, which was fun for the team because they’re so iconic. We modified a live action, practical actor. Most of our work was replace the actor’s head. The practical head didn’t look like what Chris [Carter] wanted in terms of the close-encounter alien with black eyes.
[-12:22] Allan: What’s funner: Doing aliens or doing dinosaurs?
Shawn: When you’re working on something like Jurassic World, you’re almost making a National Geographic film. There is this commitment to their being believable. If you’re looking at something like Independence Day, our ability to create a believable alien is not what the client is questioning. It’s more about the drama and the storytelling. In one case, you’re trying to nail the animalistic feel of the creature; and in another case, you’re trying to hint at its scary or its humanity factor.
[-11:05] Allan: With creature work, is there any type of work that you seek out? Obviously, you have strengths in so many areas. What kind of stuff do you get excited about?
Shawn: One of the things we try to do at the studio is try to create a diverse portfolio of work, but within that try to create a creative theme. There is so much diversity in visual effects that studios can get turned around and lose an identity for what they’re known. A lot of the core of what we do is animation. We start with animation and we build from there. So we look at those projects. The team here is excited to work on American Sniper, but all we did there was compositing. The projects that tend to excite people the most here are centered around animation.
[-09:58] Allan: That’s great! How did you end up doing creature animation work? How did you find that and establish yourself as leaders? Was it a natural progression?
Shawn: It was pretty organic but the reality of working on something like District 9 drew together a group of people that’s worked together before. And we worked on animation centered work. District 9 became the touchstone for the identity of the company.
[-08:59] Allan: That’s cool! How many of you, guys, got to touch ReBoot back in the day?
Shawn: I actually did touch it, very little of it. I worked at Mainframe Entertainment in the late 90s.
[-07:39] Allan: What about Game of Thrones? You, guys, did some massive sequences for that?
Barry: We mostly worked on Seasons 5 and 6. Most of the stuff that we did was killing people and set extensions. Which is great work, but if you’re a fan of the show, it’s traumatizing because you’re getting spoilers. We killed Jon Snow. We had to hide the work from anyone in the studio. Lots of gore! Last Season, we had to replace part of Shadrick’s face. That was a big challenge. It came late and unexpected because the client hoped the practical stuff would work. It’s what Shawn and other people were saying today: We don’t differentiate out pipeline. It allows us to do high quality work in a short amount of time.
Andrew: There was also a lot of big procedural set building for both of those Seasons.
Barry: In Season 5, when we worked on the ice wall, it was more of a Matte Painting approach. But in Season 6, we spent the time to make it work properly as an asset.
[-05:27] Allan: It’s been really great to get so much of great insight from you, guys. I did have one last question: What do you look for when artists are applying to Image Engine? Are you looking for technical ability (like scripting) or does it depend on what department they go into?
Shawn: Having the technical part is not a requirement but it’s a plus. In certain departments, you need to be more technical or more artistic.
Barry: It definitely depends on what you’re applying for and what the current balance is at the studio. The most important thing is a good fit at the studio. We look for people we want to work with and who are passionate and have similar interests with us.
Jenn: Even [during job] interviews, having a positive attitude, no big egos — that’s what makes every day so great here.
[-04:28] Allan: I’ll actually sneak this in here. It’s become a really common theme. Are there any red flags you look for when hiring people: if they’re a team player or if they’re in love with themselves?
Jenn: I think you can get a read on that. The main one for me would be: If they’re a team player.
Barry: If the artist is still hungry for the show. We do a lot of diversity: a really cool shot on one show and a really boring one on another. Sometimes the artists who are just looking to build their show reel — it makes it hard.
Andrew: You need people who are willing to mix and talk to different departments.
[-02:20] Allan: The more you understand other departments — where your work is coming from and where it’s going — the more you can anticipate what people need. Thanks for doing this! It’s been great to get insight on the work you’re doing!
Everyone: Thank you!
I had a lot of fun doing this Episode. Next Episode, I will be interviewing Freddie Wong who is a Director is one of the Founders of Rocket Jump. He does really cool stuff: a show, a podcast, etc. He’s got 4 million followers on YouTube.
In the meantime, leave a comment. I had a blast!
Don’t forget to subscribe to the show on iTunes to get automatic updates.