Hello, this is New Game Order, formerly known as ComonGames. We’ve been working on the game called The Uncertain: Light at the End since 2016. It is a sequel to our first post-apocalypse game The Uncertain: Light at the End about a world that has been conquered by AI. We will talk about building this universe in our further blogs, but now we want to begin with sharing the difficulties we stumbled upon shooting mocap.
As our team has grown to more than 20 people and our ambitions have grown with us, we knew we wanted to go further with our storytelling for The Uncertain: Light at the End. That meant creating a truly indie solution for performance capture.
Our goals are ambitious and our developers are great, but we are still a small studio that doesn't have the kind of resources AAA studios can throw at a project. Small studios have to use a lot of tricks big studios don't think about, and our greatest strength is our creativity when faced with big problems.
We worked with a very simple mocap setup, but even that simple setup made our animators' work way easier and allowed us to achieve excellent results. We are sure that many small studios have similar ambitions to ours, and we would like to share some experience in this space.
The Setup
The setup we worked with featured multiple tracking cameras, costumes with reflective markers, crashmats for actors to fall on, regular chairs, and some props that allow us to simulate surfaces in the game.
First, a preliminary calibration was done, then the in-game characters are calibrated so that their markers match the markers of the actors. Then the game characters inside our game scenes are put onto the 3D model. Of course, everything was unrealistic and without textures, but it is still more convenient to work directly with our character than with an abstract dummy.
In addition to the actors, we had a director on the set who has a general vision of the scenes with a list of animations in his head, and an animator who understands how to work with the finished material. Everyone worked together to help with props, placement, etc. We would say that this is the minimum required to make our setup work, along with some of the studio's staff who were responsible for filming.
Mistakes and advice
Large studios shoot everything: facial expressions, eye movements, lip movement and lip-synching, and sometimes even finger movements.
With that level of performance capture your actors need to play full-fledged roles like in a movie, but we didn't have this level of fidelity. Facial expressions would be hand-animated, and the voice track was recorded separately in a studio.
The end result is not perfect, but it is good enough to provide our animators with good references and achieve a better scene as a result.
As newbies to all this, we made many mistakes that will be useful to know about those who follow in our footsteps.
- Study all possible materials on the topic in as much detail as possible to become experienced yourself. Do not rely on the knowledge and experience of other people entirely, because your project can be unique in many ways.
- Prepare detailed storyboards and animation lists for everyone to see. Don’t keep ANYTHING about the capture session in your head.
- Strictly observe the nomenclature of files: names, dates, folders and files must be sorted like you are in a laboratory and keep all files on the corporate disk space in the cloud. It will save days or weeks of work later!
- It is better to use the model of the character for which you are shooting animations. Repositioning the model takes time, and sometimes it seems that it is more convenient to shoot everything in one. Still, if some confusion and the nomenclature of files are not in order, it will be almost impossible to figure it out.
- All props should be matte; otherwise, cameras will perceive glare as additional markers. Even rings and earrings can work so, so don’t forget to take them off before shooting. Use as many props as possible, it is more convenient for the actors to work, and the final result will be much better.
- Use floor markings, try to make them on the same scale as in game locations and make sure the actors follow them.
As a result of the shooting, a particular pipeline is formed: the studio gives the files with the materials, the animators watch them, choose the most successful ones, transfer the characters to the game and start polishing. Without it, animations look flat and lifeless. Finishing takes a lot of time and the final result may differ from the preliminary one quite a lot.
Not so long ago, we allowed players to play an early version of the game, and there you can see what the animation is in the intermediate stage before the final polish. You will have a chance to see how significantly better they will look in the release version. Take a look at this playthrough of Light at the End’s first level before polishing.
Unexpected benefits
What were the other advantages are there in this activity? It's fun! Yes, the actions often have to be repeated 30 times and it is hot in the studio. You have to run a lot, move the props, and in the evening it feels like you spent the day unloading the truck. But in the process is exciting and ideas are often born that could only appear on the site. Perhaps your actor will decide to replay the scene in their own unique way or someone will make a joke or a unique gesture, and the result is not what we thought of initially.
Once, we had a funny scene where our actor playing Emily pretended to be a robot, saying "kill all people", and shooting Matthew with her finger. It turned out great, and found a place for this scene in the game, although we did not initially plan it, and there are many such examples.
Want to learn more? Follow our news!