Image courtesy of Uplink | University of Hertfordshire

Uplink: an award-winning student game cinematic built with Unreal

Melissa Robinson
Students from the University of Hertfordshire recently won Game of the Year at the 2020 Rookies competition for their work on Uplink, which is an immensely impressive sci-fi short film inspired by WALL-E, Titanfall, and Portal. The development team all enrolled within the university’s 3D games art and design program, and we caught up with them to see how they created Uplink within one academic year. They talk about how Quixel Megascans assets helped them make more grounded, realistic environments and share their workflow and advice for other aspiring real-time creators.
 

Congratulations on winning Game of the Year at the 2020 Rookies competition! Can you all introduce yourself and tell us about each of your roles on the team? Have you all worked together before?

Vilte Bendziute:
Hi, my name is Vilte Bendziute! I was the main animator on Uplink. I worked with the rest of the team for the first time during my second year at the university on the Spire project. I had a good time working on the game with everyone, so I stuck around again for our final year project.

Laith Shewayish: Hi! I'm Laith, and I’m responsible for concept art, character art, and rigging on this project. We're still super stoked to have won the Rookies this year! This isn't the first time we've worked together. We also worked together on another project, Spire, in our second year, which also won Game of the Year at the Rookies in 2019.

Jamie Callow: Hi! I’m Jamie Callow, and I was responsible for the environment and prop art, lighting, camera work, and editing.
Image courtesy of Uplink | University of Hertfordshire
How did you get started on the project Uplink, and why did you choose to create a game cinematic?

Bendziute:
We wanted to make Uplink an actual game, but there weren’t enough team members for our initial idea. We had to compromise and make a cinematic trailer for the game. This gave us the opportunity to create gameplay later down the road. Also, since we had made a game last year, making a cinematic would have given us different experiences and skills.

Shewayish: We started pre-production almost immediately after finishing our second-year work. We knew we wanted to make something sci-fi from the get-go and spent the majority of the summer exploring the kind of aesthetic we wanted to go for. We initially planned to make a fully playable game demo, but setbacks early on in the project shifted us towards a cinematic as it would let us focus on delivering strong visuals.
Image courtesy of Uplink | University of Hertfordshire
Callow: In the pre-production phase of Uplink, we all tried to throw around as many ideas as we could, no matter how crazy they might be. This really allowed a lot of free-flowing ideas to be bounced back and forth. Although a lot of ideas were thrown away, this let us maintain the themes and ideas we liked among all the chaos to create Uplink.

We had previously made a gameplay demo with Spire last year and experimented with small cinematic sections within it. We realized we were really capable here, enjoyed this aspect, and really wanted to tell a short story while also focusing all of our time into a much more concise, linear experience. 

It also allowed me personally to fully utilize my time creating environments for specific shots that were much more linear and focused as opposed to a fully playable open game. Creating a cinematic also allowed me to push Unreal Engine to its limits by using the movie renderer to render higher quality frames straight from Unreal Engine. Overall, creating a cinematic really allowed us to push the engine to its limits to create a much higher quality final product.
Image courtesy of Uplink | University of Hertfordshire
What is the background story for the world of Uplink, and where did you draw your inspiration from?

Bendziute:
We were very inspired by WALL-E, but we wanted to play with contradicting characteristics a little bit. What if our robot was a massive and intimidating machine that seemed scary at first but was really a sweetheart? There was also a bit of Titanfall influence sprinkled in.
Image courtesy of Uplink | University of Hertfordshire
Shewayish: Uplink takes place on Titan, Saturn's largest moon, after an attempt at building a colony there undergoes a catastrophic collapse before it’s finished. In order to keep their failing a secret, the company responsible has reprogrammed their security drones to hunt down the colonists so nobody can spill the beans. We drew inspiration from many dystopian sci-fi works, notably WALL-E, Blade Runner, and Portal. In particular, I was inspired by the mechanical design work of Yoji Shinkawa when designing our main robot character.
What were the parameters around this project? Did you have time constraints? How long did the development take?

Bendziute:
We had our third year to create the project, from the conceptualization phase to finishing the cinematic. This meant we had nine months to create an entire short film from scratch with a core team of four.

Shewayish: We were quite fortunate that the assignment for the final major project at Herts is very unrestrictive. We had free reign to do pretty much anything! The only constraint was time. We had to finish it by the end of May. Since we started our earliest concept stages around June, it took us just under a year overall, though the bulk of the work started in September.

Could you share some of your story-boarding techniques?

Bendziute:
This involved lots of going back and forth between ideas, lots of streamlining, and compromising. We spent almost half our time on just pre-production. It took a lot of time and communication to come up with a story that everyone felt happy with.

Shewayish: We used exclusively hand-drawn animatic-style storyboarding at first, eventually blended in some live-action footage we recorded at home. I was responsible for the majority of these, and they were a lot of fun to put together! I tried to focus mainly on showing the tone and expression of each scene at this stage rather than the structure of the environments or character models. We also used more elaborate 3D blockouts later down the line.
Image courtesy of Uplink | University of Hertfordshire
Image courtesy of Uplink | University of Hertfordshire
Image courtesy of Uplink | University of Hertfordshire
Could you elaborate on how the team developed Uplink’s breathtaking environments, color palette, and lighting?

Callow: I knew going into Uplink that I wanted to push every area you progress through to feel different and keep the visuals fresh as you got further and further into the cinematic. With this in mind, at the earliest stage of the project, I made sure to spitball quick environment ideas with fast models and shapes with a bigger focus on lighting, colors, and atmosphere. Although a lot of this exploration didn't make it into the final cinematic, it all helped inform the final product due to the fact we could pick and choose what we liked from our wide selection of ideas.
Image courtesy of Uplink | University of Hertfordshire
Going back to the idea of trying to stretch as much variation in the environments as possible, I took full advantage of this and tried to inject color and lighting-value variation throughout the cinematic without it feeling tacked on. 

I added things like the fluorescent blue/purple lighting in the arcade, the harsh red alarm lighting in the control room as well as the green hologram light spilling onto the surroundings. The same idea goes for interior and exterior lighting. I made sure to have fully exterior shots, interior with heavy exterior light, completely interior shots with no exterior light, and interior locales with minimal exterior light. This all comes together to keep every section of Uplink feeling slightly new and fresh.

In general, I'm extremely into filmmaking and movies, which all influenced me heavily throughout the project, from themes to design, and most importantly, lighting. I did a lot of research into how films are lit and replicated a lot of the same techniques within Unreal Engine as opposed to lighting just using basic techniques. For example, cinematographer Roger Deakins frequently uses his own ring of warm bulbs to create a soft light that diffuses throughout the environment. I created a similar mesh and set up emissive bulbs that contribute to baking to achieve nicer bounce light when baking while also allowing me to manually fine-tune where I want this diffused light in each shot.

Another technique used in films involves box lights. I created a few pieces of simple geometry where a scalable emissive plane is housed deep inside an opened-ended cube, almost replicating a box light. When this emissive is baked into the lightmaps, it will bounce around inside this box, get heavily diffused, and leak out the end into the rest of the environment. This creates soft, indirect lighting that can be positioned wherever you want.
Image courtesy of Uplink | University of Hertfordshire
Did you use Quixel assets for your backdrops? If so, what was your experience like?

Callow:
Megascans became available for anyone to use completely for free partway through the development of Uplink. Up until this point, I had been sculpting rocks and other pieces to populate the environment, but Quixel Megascans dramatically helped to get immensely fast results and to help fill out a lot of the environment. A lot of the more natural geological assets, like rocks and terrain, were Megascan assets as well as smaller, more man-made things like cardboard boxes, debris, and trash. This helped ground the environments and fleshed them out more. 
Image courtesy of Uplink | University of Hertfordshire
How did you pull off Uplink's character animations?

Bendziute:
We originally planned to use mo-cap animation and spent quite a lot of time experimenting trying to get it to work, but in the end, we couldn’t get the results we wanted. We had to use keyframed animation instead, but we still wanted it to look realistic, so the process of animating was tons of reference and tons of feedback from the team. A few seconds of animation usually took a few weeks. We also had a lot of freelancers helping us out. Without them, the film probably wouldn't have been half as long.

Shewayish: Initially, we planned to do the majority of our animation using motion-capture, at least for the engineer character, but we had problems trying to get our captured movements cleaned up properly, so they could only be used as references. We had to assemble a massive team of animators from all years of the university to complete all of our animation work. They are fantastic people, each and every one of them. 

Can you tell us about your work pipeline? How did you organize yourself while working on the project?

Bendziute:
The whole team was organized on Discord. We had separate channels for different disciplines, and everyone could see the work that everyone else was making.

Shewayish: We were lucky enough that most of the team lived in the same household, so sharing our thoughts and ideas was pretty straight-forward, especially once lockdown was put in place. We used a Discord server for communication with the rest of our team.

Were there any tools and/or features that were particularly helpful during the production of Uplink?

Bendziute:
The fact that our film was rendered in real-time was a huge help to everyone. It meant I was able to work on animations right before the deadline, and everyone else could keep making changes to characters and environments. Not having to worry about huge render times gave us more time to polish the film before we had to submit it.

Shewayish: Unreal's robust animation retargeting features were very useful in making sure our animations kept working across multiple character iterations and ray-tracing really helped us push the visuals further than we could have otherwise. 

Did you face any challenges during production? If so, how did you solve them, and what did you learn?

Bendziute:
Everything was a challenge in its own way. For me, it was the usual problems animators tend to face during production. Usually, once you finish an animation, it’s easier for other people to then come up with better ideas or different ways to improve the scene, which will require you to re-do some of your work. The best thing to do is to make sure everyone is communicating clearly, so there is less misunderstanding later on and to not take criticism personally.
 
Shewayish: One of our biggest challenges was creating a cohesive narrative. We initially set out to have something considerably longer, but as development went on, we found that to be less and less achievable, so we were often having to rework our story somewhat to still have a satisfying start and end. While I still think it's very important to be ambitious from the get-go, make sure you have alternate paths your project can take if time isn't your friend!

What did you learn about development while working on your project?

Bendziute:
It is way harder than it looks, and it might take five times longer to finish than you originally planned. It is likely that your initial idea may greatly change by the end. It’s worth noting that having frequent deadlines is very important; if one person is behind on their work, it can easily affect everyone else.

Shewayish: It's hard! It's absolutely always a team effort and requires good management and understanding of your work pipeline, as well as constant communication between team members. Despite the hardships, though, it's always incredibly satisfying to see it all come together towards the end.
Image courtesy of Uplink | University of Hertfordshire
What are your career goals and aspirations? Have you landed any internships?

Bendziute:
I would love to animate for a blockbuster film at some point, or for a popular TV show. I also like the idea of being an animation director one day. I recently had the opportunity to work with Rewind as a freelancer, which was my first experience animating for a big, well-established company, and I had a good time. Everyone I worked with was super nice.

Shewayish: I always wanted to be a concept artist since drawing and painting are where I have the most fun, though 3D is great as well! Luckily, I started a job as a concept artist at Frontier Developments this week, which is a real dream come true. I definitely owe that to this project, and our Rookies win.

What tips would you give other students who are starting out with Unreal?

Shewayish:
Learn the basics of Blueprints! Even if you don't plan on being a developer or tech artist, having a basic understanding of it can help make your work more cohesive and interactive, and help you integrate your work more easily with others. Definitely play around with those new ray-tracing features, too, if your PC can handle it. It can make a big difference. 
Image courtesy of Uplink | University of Hertfordshire

    Are you prepared for the jobs of the future?

    Sign up for the Education Newsletter to receive monthly updates with free Unreal Engine learning resources along with other helpful tips for navigating the exciting world of interactive 3D.