By Virtuos Games
Games these days aren’t just about earning points and finishing levels anymore. In just a few short decades, games have progressed from Pac-Man and Frogger to masterpieces like The Last of Us and God of War, each of them offering a fantastic experience comprising exciting gameplay mechanics, jaw-dropping scenery and an engaging tale to boot.
The emergence of storytelling in games has led developers to borrow ideas from the film industry, prompting the use of in-game cinematics as a tool; both to act as checkpoints marking significant milestones for the player, and to progress the story being told within the game.
When it comes to crafting cinematics for games, Sylvain Doreau knows best. As the newest Executive Creative Director in Virtuos, Sylvain has had an illustrative career spanning decades in both the film and games industries. This has naturally led him towards many unique opportunities, including the chance to collaborate with the likes of George Lucas at ILM and the talented people in Activision, working on some of the biggest media franchises in the world. Star Wars; Call of Duty – he’s done them all.
With a history such as his, it would be unthinkable not to ask him out for a chat. Here follows the complete interview with Sylvain, where we delve into topics such as the nuances of directing a game cinematic, how the future of making cinematics will look like, and his passion for rugby outside of work.
Thank you for joining us, Sylvain. Let’s start with a description of your career history in brief, leading up to Virtuos.
A brief one will be quite difficult, but I will try my best!
After finishing film school, I started working as a freelance CG Artist in Paris in the early 1990s, before starting a CG animation studio with a friend in a post-production house. That quickly became Medialab Paris, owned by Canal+. We were the first to do real time animation and motion capture in Europe, working on CG feature animation projects and animated television series.
But France was not ready for feature animation yet at that time. I left in 1996 for California, where I was hired to be a part of PDI-Dreamworks Animation. I was fortunate enough to start the adventure at the very beginnings of the studio and work on two great CG animated movies: Antz first, then Shrek, which proceeded to win the first Oscar for Best Animated Feature. I also started what we called the ‘rough layout’ department there, which is what we call ‘previsualization’ or ‘previs’ now.
In 2001, I had a great opportunity to enter the games industry, which I did by moving a few miles north to join EA Redwood Shores (now Visceral Games). I started directing cinematics, scripting game cameras for several franchises, and editing cinematics and trailers. This lasted until 2006, when I received a call from a friend at Industrial Light & Magic, about working with ILM’s R&D department to help develop a real time previs tool.
A few months after that, I found myself working with George Lucas and Steven Spielberg in person on the previs for Indiana Jones and the Kingdom of the Crystal Skull. This granted me entrance to Skywalker Ranch, where I started up the 3D previs pipeline and creative process on the first seasons of the Clone Wars TV series. I also worked on the feature animation pilot for Strange Magic with George Lucas directly for a few months – a somewhat stressful but very interesting experience!
My Lucasfilm adventure concluded at LucasArts, the games division back in San Francisco. There, I directed the cinematics for Star Wars: The Force Unleashed II and dealt with in-game cinematics and cameras for several other games.
In 2010, I got hired by Activision and moved to the headquarters in Santa Monica, to lead and expand their cinematics capabilities across all franchises. I was overseeing the creative side of things, while directing more than 100 cinematics for franchises like Skylanders & Call of Duty among others. During my tenure, I collaborated with all the worldwide development studios, and was in charge of selecting the cinematics partners and external CG animation studios. I also dealt a lot with cinematics production and the dev studio/partner relations that were sometimes challenging. I’ve probably never travelled that much for work in my life as I did back then!
In 2017 I started to do consulting work in Los Angeles, making games trailers, commercials and cinematics with my production company Spacecargo and ultimately moved to Ireland in mid-2018, to help Hasbro’s animation division jumpstart their feature animation studio in Dublin.
What do you think is your biggest achievement to date within the games industry?
I’ve worked on AAA games before at EA and LucasArts, but with Activision, we had the opportunity to contribute every year to mega franchises like Call of Duty, Skylanders and Destiny, which was truly a blast. Every release was massive; their marketing and sales departments are like war machines, and the months building up to each release were always incredibly exciting. You couldn’t open up a web browser or turn on the TV without seeing a banner or trailer about it, there were giant billboards everywhere in Los Angeles, and so on. Being so busy working on these juggernaut games for many long months, you don’t really realize the expectations, the hype and excitement they create for the fans and press until you finally look up from your desk.
There is one particular thing that I remember very well. We were in New York with the Skylanders creative team, right after the release of Skylanders: Giants and already working with our cinematics partner on what would be next year’s project (Skylanders: Swap Force). On a lark, we decided to take a look inside the Toys “R” Us store at Times Square. There, right at the main lobby, was a GIANT three-meter high statue of Tree Rex (one of the characters in Skylanders), and our games were placed everywhere! No one would be able to miss it. I was like, “Wow, this is amazing.” I couldn’t believe that this is what we do for a living!
The dark and abstract cinematics we did for Call of Duty: Ghosts were a big achievement, but the one project I’m exceptionally proud of is the Exo Zombies DLC cinematics for Call of Duty: Advanced Warfare. There was less pressure to produce the DLCs than the mainline Call of Duty games, but there were decent budgets for cinematics still.
We had the opportunity and budget to choose an amazing cast of zombie fighters: John Malkovich, Bruce Campbell, the late and super friendly Bill Paxton, the ‘Punisher’ Jon Bernthal and the iconic Rose McGowan. We scanned their faces, recorded their voices in locations around the world, wherever they were, and the motion capture sessions were great fun.
Working with these legendary actors was a once-in-a-lifetime experience, and the final pre-rendered cinematics still looks amazing to this day. I carried my Leica camera through all that process and actually did a ‘Making-Of’ book for the team.
So, what inspired you to join Virtuos?
I don’t think there’s another organization like Virtuos that covers that much ground in terms of game services and has an equally broad set of skills and talent. Starting a new business unit, building teams, and working with existing studios and partners around the world is something I really loved to do in the past, and wanted to get back to.
I had the opportunity, thanks to Olivier Masclef at Black Shamrock, to pitch the idea of a cinematics group to (Virtuos CEO) Gilles Langourieux last fall and met him in Dublin earlier this year – and here we are now! I got really attracted by the notion of end-to-end services and the concept of co-development. The notion of ‘creative development’ was something that was missing in our services, and I’m proud to say that we can truly offer end-to-end cinematics services now!
Something as important as the job itself is the human factor that I’ve found within Virtuos. When I first began interacting with Gilles and the management team, I was amazed to discover the fantastic diversity and cultures you can find at Virtuos. Having studios around the world and working with people from different cultures, whose first languages may not necessarily be the same as yours is something truly special. Having lived abroad for 20 years, I’m a kind of multi-cultural chameleon and I feel at ease acting as a bridge between different cultures at work, while understanding and respecting people’s differences. For me, it has to be a rewarding human adventure and experience first. Unfortunately we can’t travel to meet the teams at the other studios yet; Zoom and Skype are great, but I’m looking forward to interact with people for real!
As Executive Creative Director at Virtuos, what do you hope to achieve in your new role?
Like I mentioned earlier, expanding our end-to-end game services to include cinematics is particularly essential. The addition of creative development capabilities is a big plus: we will be able to offer services ranging from concept, storyboard, directors, to writers, mocap sourcing and previs. Adding creative services is something that will put us on the map as a global studio – not only a ‘cinematic service only’ studio, but now a full creative services studio. That’s something I’m going to really push and develop.
Another advantage for Virtuos is that we already have a network of clients, and we’re already embedded into their development teams. Cinematics is therefore not a ‘new’ service, but ‘more’ service! Being a one-stop shop is a big plus and will prove to be very attractive for our existing clients.
Aside from expanding our scope of creative services, we’ll also need to remain competitive in terms of tech, pipelines and skills, which is why we’re going to try to push for more in-game cinematics to create seamless next gen experiences. That’s going to be our main direction, so developing solid Unreal and Unity pipelines is critical. Sparx* is already using Unreal as a go-to tool to assemble scenes – doing matte paintings for example – regardless if it’s pre-rendered or in engine content, and we’ll extend that to previs shortly.
Pre-rendered cinematics and to some extent, high-end game trailers, still comprises a big part of the market, so we’ll still concentrate on being creatively and technically competitive in that particular segment. We’ll need to find the right partners and develop internal skills as well.
One last direction I’d like to push towards – a stretch goal, if you will – is game-based linear content. Our big advantage is our TV animated series experience at Sparx; I believe it could really help us to create in-game animated content (real time or pre-rendered), TV series, animated shorts, web-based content – things like that. That’d be a great way to leverage on Sparx*’s skills and create animated content that is 100% game related.
In your opinion, what defines the term ‘narrative content’ in the context of games?
Narrative content can essentially be defined as a series of story based sequences, done either in engine or pre-rendered CG, for the purpose of exposition. This usually translates to non-interactive content or scenes that may include several shots or cuts, which will inform the player about a particular aspect of the story or gameplay. Things like character exposition, an intro to a new boss, a new game hub or level, a puzzle or a mission brief, which helps the player to progress through the story and the game.
The thing about narrative content is that it used to be considered as ‘non interactive’ – boxed, isolated, separated from gameplay, which means it’s generally done separately from actual game development. It also requires large teams and is often created by external studios that have the capacity to handle large amounts of high quality CG animation. So technically, the cinematics fall in that category.
With that said, cutscenes that feature some element of interaction are becoming increasingly common, particularly for in-engine based cutscenes. One example is the ‘button-mashing moments’ that have become a staple of the narrative content. Games like The Last of Us, God of War or Uncharted for example have lots of these.
Some games – especially Quantic Dream’s ‘story-heavy’ games like Heavy Rain and Detroit: Become Human – feature these as the main mechanic of their games. What do you think of having interactive cutscenes represent the majority of a game?
Having a games based on interactive cutscenes creates different challenges for the narrative and game designers, particularly in terms of writing and flow of the game. Games demanding extensive player input tend to be longer and slower in pace. Detroit is certainly an example as you’ve mentioned, it’s also the case for games like the Mass Effect series. When you start to introduce multiple choices in your story that will influence the outcome of the game, (Red Dead Redemption being yet another good example), the narrative structure will become a giant arborescent network, with a play time easily entering the double digits.
Meanwhile on the opposite side of the spectrum, you have fast-paced first-person shooters that usually feature a shorter playing time in the singleplayer campaign mode (about 6 to 8 hours), with more traditional non interactive cutscenes, generally at the start of each level, mixed with fewer interactive-in engine cutscenes during the course of the level.
How important do you think narrative content is for games? Should all games include it whenever possible?
In today’s games, the line separating narrative and game interactive content can be very blurry. The story is essentially embedded into the gameplay and we see less the format that used to be the norm: a pre-rendered cutscene (that the player would want to skip after the first time), the mission brief, then gameplay, then back to another pre-rendered cutscene.
In contrast, today’s narrative content tends to bleed into the game content. The PS4 and Xbox One are powerful, but still a bit limited in terms of disk space and performance to offer a truly seamless experience between gameplay and the ‘story-based content’. The next gen consoles, with their capabilities, will enable even more seamless experiences, where narrative and interactive/gameplay content will intertwine gracefully. This also means that narrative directors and game designers will no longer have to limit themselves to older designs and story patterns, but will be able to explore more innovative forms of interactive storytelling. For a hint of what’s to come, we can look at games like Red Dead Redemption, where your choices will profoundly affect the narrative content or a masterpiece like The Last of Us, with the story constantly being embedded into the gameplay.
There’s also a trend to include more and shorter ‘Cinematic and scripted moments’ as player rewards and mini events, which could also be considered as narrative content (the latest Mortal Kombat games have lots of great ones). There’s dialogues, authored cameras and custom animations, which are all basically mini narrative moments designed to enhance a gameplay moment. It could be minimal in terms of story content (i.e. ‘You die, I win, I progress to the next level’), but still performing a narrative function.
Another big trend we’re beginning to see more is the emergence of game-based linear narrative content. Films are included in this category of course (while not always the best examples…), but for us specifically: TV and streaming animated series, short animated CG films, or other web-based content. Taking League of Legends as an example, one can now expand the player’s experience outside of the game itself, into a narrative format that allows for more backstories, going in-depth with the game’s characters and the universe in general. Riot Games has done some amazing work with their TV series, shorts and narrative trailers, to compensate for the lack of narrative content in the actual gameplay. It keeps their players engaged, even while they’re not holding their mouse or controller! That’s something we also should approach and be good at, especially now that we can handle creative development.
Is there a difference between making game cinematics and fully animated feature films?
Yes there is. First, the story structures are vastly different for animated films and games; in films, you generally have a traditional three-act story structure, for a running time of approximately 70 to 90 minutes.
On the other hand, games can be as long as 40 hours or more for certain titles. Some game genres, like MMOs, more or less never end! The games that would be closest to film in terms of story structure are typically linear story-based games, like the Call of Duty games or action-adventure games like Uncharted, for example. They are going to have somewhere like 6 to 8 hours of gameplay – sometimes more – and about 30 to 45 minutes of narrative cutscenes. Generally, the story is not open-ended and goes linearly from beginning to end. Open world games are generally completely different and the multiple quests and such make the story structure way more complex than Toy Story or Shrek!
What about the production side of things? How different is it to make an animated film compared to a game cinematic?
There are similarities and differences depending on each stage of production; I’ll try to explain by going through them one at a time.
Let’s start with the initial design process, which is largely similar across both formats; once you have a story, regardless of length, the Art department creates concepts, character designs, key art and color scripts.
Things start to differ when it comes to storyboarding. In feature animation, you tend to storyboard everything; every sequence is done multiple times before the project even starts. For game cinematics however, it really depends of the type and genre of the game. If you were to do a linear story-based game for example, you would want to storyboard your sequences before going into production – basically using placeholders early on in the build to check the pacing and how it flows with the game. But for open-world games, some action-adventure games and mocap/performance capture-based projects, lots of narrative directors and cinematics directors love to keep some room for improvisation and let the actors experiment during the mocap shoots, which can be a double-edged sword. If you’re really in control, you can get great results, but it’s also possible to end up with a massive number of mocap takes that will require time to stitch everything together in editorial and mocap assembly. Doing mocap can also be dangerous, expensive and finish with clunky results. Having actors perform with no idea where the camera is going to be (it’s generally applied afterwards) could lead to weird eye direction and continuity issues.
The previs phases could be quite similar between CG cinematics and CG animated movies in that they mostly use Maya, but more and more cinematics previs and sequences are directly assembled in real-time engines like Unreal or Unity these days. It’s a way more flexible process and you can explore more options without increasing your cost. Cinematics cinematography uses the same composition principles, but again, real-time engines are now commonly used to tune cameras. In addition to that, virtual cameras can also help create a massive amount of shots in a single day, compared to the traditional Maya-based feature pipelines.
Keyframe animation is more or less the same with film and games, whether you’re dealing with something from Skylanders, Ratchet & Clank, or a Rancor from Star Wars, and you’ll similarly need big teams to create 25+ minutes of content no matter the medium.
Mocap and Performance capture-based animations are commonly used for in game and pre-rendered cinematics when you deal with humanoids; mocap understandably works better for realistic characters than for cartoony animation. The lighting and FX pre-rendered pipelines are similar for feature animations and pre-rendered cinematics, using more or less the same renderers (Arnold), and Houdini to create explosions and simulations, with compositing done in Nuke.
For in-engine game cinematics however, things are completely different, done entirely in game engines (Unreal, Unity). There’s no ‘compositing’ here per se, but we do still think about stuff like ‘post-FX’, vignettes, depth of field, flares, color correction, all of which that will give a similar look as a pre-rendered cinematic without hours of compositing.
Performances and frame rate will be a huge issue at that stage if it’s all handled in real-time. Often, the real-time cinematics are ‘post-rendered’ as a single frame and exported as a compressed movie in-game (or used for marketing purposes). But with projects like The Mandalorian which we had contributed to, we’re starting to see real-time engines (Unreal in this case) being used to do real-time compositing for VFX projects. Will feature animation follow that trend? Potentially yes, partially to start with (for previs, backgrounds and such), but it will definitively evolve within the next few years!
Interesting that you mentioned the increasing reliance on engines like Unreal to do real-time compositing, because Unreal Engine 5 was just revealed last month that claims to have huge improvements on that front. Any opinions, predictions or hopes about Unreal Engine 5 in particular and what it can do for cinematics production in the future?
Unreal has always been doing amazing reveal demos. It generally includes insanely complex environments, with gazillions of polygons and great real-time rendering. While we’re starting to see the VFX industry relying even more on games engines, cinematics will also potentially use less traditional CG animation and VFX pipelines at the same time. Cinematics will be done mostly in-engine, even if they’re ‘post-rendered’ for compositing in post-production. We’re already using Unreal to assemble scenes at Sparx*, even for pre-rendered projects, and it’s going to become the new norm. The production pipelines will be mostly based on real-time engines, while traditional software like Maya will only be used for specific tasks like animation. The overall game experience will be visually more consistent and you’ll move seamlessly from gameplay to cutscenes and back again.
With all that said, one thing to be cautious about is that in these great demos, you never see more than 8 to 10 characters with full clothing and hair simulations and full facial features, interacting together in the same scene, at that level of quality. Rendering tons of rocks in a canyon environment seems to be ‘easily’ achievable today, but the success of the next-gen engines with cinematics will heavily depend on their ability to digest and process a vast amount of character based animations and simulations in real time. We’ll see what the next two years will bring us, but our goal is to mainly work in engine and push the medium’s boundaries.
In your opinion, what’s the most important part of the process when it comes to crafting game cinematics?
It’s tricky to single out a particular part of the process. In my opinion, there’s more than one that’s ‘most important’ – but regardless of the type of cinematics or client, the story is critical. A well written and well-paced and timed script, with clear descriptions, good dialogues, and meaningful characters is key, and it forms the foundation for a good cinematic. Then you’ll need a well-timed previs or animatic with great cinematography, elaborate staging, character blocking and meaningful shots, which are also very important to start with, particularly if you do pre-rendered Cinematics.
If the project involves the use of performance capture and mocap, the acting direction is of course incredibly important. Giving context to actors, building props they can climb on and interact with, showing them game art, concepts, storyboards is key; knowing when to overact (action scenes generally) or be more subtle, in dialogue or important story moments makes a big difference. As a former director of photography, I’m also particularly picky when it comes to cameras, lenses choices, and composition. I love trying to reproduce real world gear like a steady cam, or Movi rigs. The audience needs to feel that the camera is ‘operated’. Whether it’s with in-game cameras, keyframed animated cameras, or virtual cameras operated in mocap stages. It helps to sell realism, as well as meaningful lighting and VFX of course.
Working hands-on with the design and assets team is also key when dealing with in-engine cinematics, as you are totally dependent on game design and game assets. Any design or asset change could have a big impact, especially in the later phases of production on cutscenes. Besides a strong creative leadership, Production has a huge role to play in maintaining communication and clear workflows between the game and the narrative/cinematics side and the client.
Which leads to the final ‘most important’ point – creating game cinematics is a generally a collaborative process, but they need to be ‘directed’. You need a vision, a clear path. Sometimes you have a client that has a strong cinematics or narrative director, who know what they’re doing – they might push you out of your comfort zone and it might not be your vision that you’re working on, but you’ll be fulfilling theirs. As long as you accept and respect it and do your best, it’ll work out just fine.
There are also other projects where the client doesn’t really have a ‘director’; you will have to face a committee of producers, often execs, sometimes, AD’s and you’ll have to filter a lot of opinions into real action items for your teams. In a nutshell, you’ll have to direct and lead the creative part, while also being a master diplomat and a bit of a producer. Both of which are plenty challenging on their own already!
Let’s switch gears to more casual stuff. Where are you currently based in? How do you like the place so far?
I’ve been based in Dublin, Ireland for two years. Coming from the giant sprawl of Los Angeles, we were looking for a city built for humans, not cars, where people have real interactions and love to take time to chat to strangers. Believe it or not, after 7 years in LA, we missed the rain and getting to experience the four seasons! Aside from that, we wanted to get away from the extreme weather and traffic, so moving to Ireland was perfect.
Dublin is a wonderfully nice city with a very manageable scale – “the smallest capital in Europe”, as they say here. There’s plenty of bike lanes to share, which means we actually don’t own a car, something that’s unthinkable for our American friends! People are absolutely fantastic in Ireland. They hate to show off, aren’t too brash and they LOVE talking, which is great news for a talker like me! We really felt welcomed and integrated very quickly into the community, and also found a great school system.
Believe it or not, the weather in Dublin is OK. Ever-changing and wetter for sure, but it doesn’t bother us. And of course Ireland is such a gorgeous country; just 25 minutes away from the city center and you could be by the ocean, in a forest or up in the Dublin mountains. It just takes a few hours to cross the island and be in very remote and breath-taking places, enough to make you feel like you’re in a Game of Thrones episode. It’s also a great hub to travel to Europe as it’s close to everything, and of course the endless conversation with strangers at the pub are unique.
Tell us a bit about your personal life. How did you get exposed to technology and games?
As a kid and young teenager, I was very much into the ‘classic’ arcade games, like Space Invaders, Asteroids, Pac-Man and such. In France you can go to cafés at a very young age (13-14), where I would spend a lot of time with friends trying to beat the Space Invaders scores or playing Pinball at the back of the cafés. It was noisy and great fun, perfect for a teenager.
But the first Star Wars movie – A New Hope had a major influence on my career. I saw it at an outdoor cinema in the summer in Provence, under a starry sky – that was magic. The storytelling, the visuals, I’d never seen anything like that. We went 3 nights in a row! One shot in particular did struck me, which was the CGI shot of the Death Star hologram at the Rebel base. I saw that and said to myself, ‘wow, that’s what I want to do!’
I had no idea how, but I managed to go to a high school that had an intensive art program. There were no computers in schools yet, but at barely 17 years of age, I had the opportunity to meet a group of former students who had just started using computers to create art, scanning drawings through medical scanners and using CAD-CAM program to do sculpture. I joined their art group, and that was it: I was hooked!
But no one was really interested yet in CG animation in the mid 80’s and there was no training or art school programs relevant to it. It was still a very new craft, and traditional animation schools were very skeptical of the whole thing (hopefully things have changed now). So after my baccalaureate, I joined a film school in Paris to learn the trade of filmmaking, and while trying to figure out how to work in CG along the way! I was a director of photography in my team (everyone wanted to be the director or writer!), which was a very technically challenging role at the time. It was mostly on film, with no video assist and no undo, which meant no mistakes tolerated. I had to dive deep into the ins and outs of film and camera tech, lighting, and also video and editing, cutting ‘real’ film in the good old days. While still at school, I managed to get jobs in various TV and post production studios in Paris to do TV graphics and CG animation. People were desperate to find artists that had some basic knowledge of computers. I basically had to learn on the job by looking at the documentation. It was tough, but it really helped me to build a solid foundation in post-production at a very early age.
After graduating from film school I freelanced for a couple of years in France, then met a genius and very ambitious entrepreneur (Alain Guiot) who already had the biggest post-production studio in Paris. We started a CG animation department that grew from 2 passionate friends into a 90-employee outfit within three years. From traditional CG, we started to do real time animation on huge and very expensive SGI workstations only used by ILM at the time. We were also the first to do real time animation for TV, such as virtual anchors and virtual sets. It’s personally interesting to see that virtual production is now making a comeback over 25 years later, and is currently one of the hottest topics in the industry!
Canal+ bought us in 1993 and we had full license to experiment on basically anything we wanted and put it on TV – truly good times! We also started to work on the one of the first feature animation projects with the sci-fi comics legend ‘Moebius’, and developed one of the first feature animation pipelines, using our in-house real-time engine for facial and body animation (which was also a first).
The real time engine was called P.O.R.C (Puppets Orchestrated in Real-time by Computer).
But we were far from the Mandalorian real time pipelines and CG still had to be rendered. Render times were really long at the time, even with computers the size of refrigerators, and so the team had lots of spare time to play video games! We loved the first Doom games, Tetris, Rayman, Twisted Metal on PS1, Super Mario on Nintendo, and Panzer Dragoon on Sega Saturn. Warcraft 2 was also quite popular with the artists, as I remember.
Those were very exciting times for the CG animation industry in Europe. I could have stayed in France, but there was something on the back of my mind that convinced me to move to California and live new adventures, and maybe one day work at Lucasfilm and on a Star Wars project to fulfil a lifelong dream. I had a few friends already working in the industry over there, and so made the jump in late 1996 to join PDI Dreamworks that was starting their animation studio in Palo Alto in the Silicon Valley.
