Gerben Pasjes - Portfolio

Intro
Keep Dreaming was our first game with a custom built engine and tooling. Many technical restrictions for polycount, texture resolution and maps made the art team get very creative with their budget. It was during this game that I again felt I had stepped up my game with concept art, making extremely useful paintings and designs that helped everyone on the team.


photo

Visuals
We fairly quickly settled on making a 3rd person dungeon crawler, with a touch of procedural generation. Our brief stated the game had to be using modular 3D meshes, so as concept- and environment artist I made a set of paintings that displayed the same environment, but using a different theme set and a few swapped models. Our modeler then used those paintings to create the full set of modular building blocks, and I created the textures for all those blocks in all four environment themes.


photo

A lot of research went into how the surroundings' values were compared to the characters and how effect- and actionheavy games like Diablo 3 solved that. We decided to use a black outline around all the assets with which you can interact in the gamespace. This was an especially elegant solution because it worked well with the pixel art style we had in mind for the assets. For the environment we decided pixel art textures would do the job nicely because of texture budget and artstyle.


Content creation
As said before, all of our content had to be created using custom built tools. One of our talented programmers built editors for the team to build levels, gear, consumables, npc's and dialogues. For an asset to be in-game, I used a collection of editors to select a mesh, a texture for each theme, a tile icon, place that tile in a level, place the level inside the world, and launch the game. The tools enabled and encouraged thorough collaboration of every team member. Definitely a good experience to be closely connected to every aspect of the game.


photo

Review
Keep Dreaming turned out to be a reasonably fun game but an astonishing technical achievement. Eight weeks of development time was enough to set up the framework for creating engaging content, but sadly it was not enough to also build and publish that content.