Rebearth is the second game project I worked on at Future Games. It was a four week project in Unreal Engine 4 with the themes adventure/exploration, no word narrative story telling and destruction. The team was 4 game designers, 4 3D artists and 2 2D artists. I worked closely with the 2D artists and another level designer to create a narrative, which was eventually mostly dropped due to sickness in the team (I’m never sick though!). Then, I worked on puzzle and level design with the scripters and 3D artists.
Rebearth is a first person puzzle game, where your mission is to help a dormant race on an alien planet to wake up again. You do this by finding orbs which step by step brings the spring to the planet. The orbs are hidden past puzzles of blocks, which you have to move around to create a path for yourself, a path that leads to the orbs.
This project was a big lesson in overscoping. I initially wanted to have nine levels, three for one axis, three for 2 axes and three for all axes. I made six levels, two of each. But only three levels ended up in the game. While I was fast in blocking out levels and making puzzles, there simply wasn’t enough time to art the levels to make sure they look good. That takes a lot more time, and when you have to consider things like lighting, a tight schedule and sickness, it’s just impossible. Still, I’m happy with the result. Because I managed to keep the three best levels, and they have a nice progression of the difficulty. Our team got the highest grade for Rebearth, a fine feather in the cap.
The main mechanic of the game worked like this: you rotated platforms 90 degrees, either clockwise or anti-clockwise. As you progressed in the game, you could do it on more than one axis. But here, on the first level which I show below, you could only do it on the Z axis. With the level, I wanted to introduce the mechanic with its limitations and to have an easy difficulty. I aimed to guide the player with light and plants showing the way.
I began the level design by making paper prototypes. There were two reasons. In the beginning of the development, there were no blueprint scripts made in the engine yet. And it is fast and easy to do. For a puzzle like this, testing its functionality works perfectly. I drew a bit on a paper, made some cut-outs and tested by moving the cut-outs around until I had a working design.
White box block-out
After the system was set up in the engine, I began blocking out the puzzles in it. Here, I could iterate on the design, to make sure the difficulty and pacing felt right. Meanwhile, I planned the graphic assets with the 3D artists to make sure we had what we needed for the puzzles.
Once the graphics were made, the blocks were replaced. That is the easy and fast part. The tougher part is set dressing the level. It took a lot of time to place the assets so they look good together and to make sure the player understands where to go and can take that path easily, yet still being challenged by the puzzle. In the end, we have a level which is pretty to look at and tricky to play.