As you may be aware, Prisoner of War has been in development for about 18 months now. Due to the direction the project has taken, towards a more narrative-led game, the cut-scenes have become an increasingly important part of the game. That's where myself and my colleague Moog come in, as it is our job to produce a series of cinematic sequences to tie the game's plot together. As we both joined the project towards the end of the development cycle, it has obviously been an immense challenge to become acquainted with the game engine and toolset.
Prisoner of War's cut-scene system is, at its simplest, a sequence of real-time game events that are triggered at specific times. These events can be directions for the in-game characters; manipulation of objects; placing cameras and cuts; setting the time of day and the weather; and just about anything else that can happen in the game. Each cut-scene is comprised of a sequence of discrete chunks of time - known as scenes. As we are using the real time game engine, and the frame rate can vary, these scenes have their duration in seconds rather than frames. This concept of scenes is loosely comparable to key-frames.
The tool-set we use really shows the quality of the Atlas Engine - the technology engineers here at Wide have done everything they can to hand control of the game content into the designers' hands. This has really empowered us to try out a number of different things, without having to constantly harass programmers for changes. The tools are so powerful that literally anything the game engine can do is available to us through a set of simple buttons and dialog boxes.
A visitor from the film industry saw this tool set running, and couldn't believe they didn't have this kind of thing when creating real films - as it allows shots and events not possible in the real world without huge expense (for example a helicam is a simple click of a button in Atlas - but a �Ks a day expense in the real world) - so again allows the film creators to try things out very quickly without wasting film, etc.
It is also useful that the cut-scene tool set fits in with all the other tools in the game - allowing us to work closely with the other designers and change elements of the game to fit with the cut scenes, all quickly and without impact on other members of our team.
However, the process of creating cut-scenes can be quite time-consuming. As with most things, the cut-scenes are usually mocked up first on paper as storyboards. The next stage is to set up the characters and locations as required, and act as a kind of virtual director, ordering around the characters and positioning the camera. In theory this allows us to work without having to worry about things like animation and collision detection; we just give orders and allow the characters' AI to take care of everything else. Inevitably, this kind of abstraction causes its own problems; even the most advanced AI characters don't always behave themselves!
We are working in a dynamic environment where things change from day to day. New features are added to the game; the mission structure evolves and scripts are revised to reflect this. And, at the time of writing the voice actors have yet to be recorded. Because our foundations are continually shifting in this manner, there's always the possibility that certain sequences will need to be re-worked. I have been looking into the possibility of using Maya or 3D Studio Max for the bulk of the work, with the required data exported to the game engine when ready. Unfortunately, Prisoner of War has come too far to develop and integrate this technology; with deadlines looming everybody's main concern is just to get the game done on time. But understanding the system's strengths and limitations has, I think, allowed a certain unique style to emerge. This project has been an important learning process for the team and an equally important development stage for the tools. Hopefully, the final quality of cut-scenes will reflect our efforts, and provide invaluable feedback for the future.