Jaren Turnbow

This was a graphical test with TF2s assets. Here RTX is being used for reflections. This game was a pet project to see if I could port source engine TF2 into Unreal Engine.

I took the opportunity of translating a movement shooter to vr, by attempting to make all sorts of new control schemes. This one is a scheme where your directional input, is based on where your head is. It's a strange one, but its surprisingly immersive. It made movement feel a lot more real. Though yes, less accurate.

This was a network test for two AI teams going at it. One pushing the payload, the other attempting to stop it. For basic AI, they actually ran really well. But you can also see plenty of player mechanics are in place too. Health and ammo kit spawners were eventually populated around the map, the resupply locker was added and programmed, jumping off the map will kill you, and a few other things. I worked very hard on making it high performance. I didn't want a TF2 fan project that ran worse than the original old game. So I worked up to giving this game a framerate of 900fps+. All while in original TF2, I had a framerate of 200. With how well I had it running, I theorize you could have made it run on android.

This was to see if it was possible to allow full Quest 2 hand tracking be the controls for a shooter game. Where your hands position was a joystick, and to allow you to make finger guns to actually control guns. It was really cool, but I don't think it's practical against actual controller players.

Having plenty of experience mapping for the original Source engine Team Fortress 2, I was most certainly underwhelmed by the level designers toolkit. A large upset were the track path tools for the payload gamemode. There were in fact none, well, virtually. The track tools were no more than a generic entity you can place in trails, and have them each reference their next and previous track entity. It was messy, easy to break, and difficult to match against the 3D track models layed out.

I put in the effort to ensure lining it up to the track models were as easy as possible, so I had built in wheel visualizers that populate themselves at every spline point along the track tools. Along the tracks are checkpoints for the cart to reach, and be unable to be rolled back farther from to ensure the pushing teams progress is kept to a certain amount. This was handled by having a separate track between each checkpoint, which was quite easy to do. Yet even so, each track can still reference it's next and previous in-case the cart should need to roll-back. But otherwise, each track has a programmed in option to make it's end a checkpoint, win, or special scripted event.