New this week

This week I created the first 2×2 room greybox and worked on blend trees. Time spent was cut a little bit short because of Christmas and the surrounding commitments but I believe good progress was made.

Last tile for a while

The last basic greybox tile I’ll work on for a while is the 2×2. Larger tiles are far easier to design than the 1×1. A lot less time was spent worrying about ensuring the doors couldn’t be seen from each other. This is because filling out the levels itself naturally closes off the line of sights from each other. The larger base tile size also allows for more gameplay diversity in the tile itself. Within this greyboxed tile there is a large open area, which is the 2^nd elevated level. The ground level features more corridors and places to kite. Another benefit of this tile is there was more of an aesthetic vision going into it so imaging this tile will look quite good when we eventually get to the decoration phase.

Blend Trees

I was expecting blend trees to be far more complicated than they turned out to be. Theres still some issues I’ll need to investigate but for an initial implementation it seems good. The process involves converting the navmesh’s desired velocity into a relative velocity then passing the x and y into 2 floats on the animator. A free unity asset used for testing which contains a rigged model and animations was found and passed into a blend tree. The parameters for the blend tree come from the velocity values. One issue with the Mecanim is blend trees are not save-able assets by default but this was remedied with an editor script. The reason blend trees need to be saved is because of the way my state machines are set up. Because my state machines are built off the animation system, I need to use the same blend tree in multiple states. Re-creating a blend tree for each state it is used is not only annoying but probably inefficient.

The issues I’m facing are I haven’t implemented foot IK, and some animations slide when direction changes. Hopefully foot IK solves both these problems.

Whats next?

Next is to work on the model and rigging for the basic enemy, the barrel kin. I imagine this will take up most of the week and if there’s additional time, I may start animation. I am functionally inexperienced with animation so that will take some time. Especially since I must get animations that loop and blend well together. Maybe if I don’t want to do that yet I will look into foot IK and see if I can get some of the animations looking better.