Not a huge improvement here, but what do you expect in a day? Got a new box model from the artists and I learned the incredibly cheap trick of setting a base global ambient value to simulate some of the effects of a global illumination system within a local illumination system. As basic a trick as it is, it’s still very effective.
This image was taken on 9/8/12.
GoodGraphics10.png, global ambient is neat.
This wasn’t much of an update from the previous post. In fact, it was even taken on the same day. After getting the basic lighting and normal mapping in, I looked at the asset loading pipeline and reworked it to easily allow the loading of different models at once. A simple task, but one which I hadn’t really considered carefully enough during my initial system design. A small refactor later and everything was ok.
This image was taken on 9/7/12.
GoodGraphics09.png, a cube of different models.
With summer school over, I took a small break from doing anything to recharge. I’m usually pretty unrelenting about working, but sophomore year had been especially draining. After that break, I got started on implementing a basic Phong based lighting model.
The two major improvements here are a single directional light used for illumination and normal mapping on the skull model. The illumination uses no global ambient to prevent the scene from being nearly pitch black, which makes for a really excellent screen capture. Also, the process by which I read in normals from the mesh file was undoing the work the artists had done in applying smoothing groups to the model, thus the incredibly blocky nature displayed here. For as much of a step forward as this was, it was also a very primitive step.
This image was taken on 9/7/12.
GoodGraphics08.png, basic lighting and normal mapping.
There’s not really any updates on the graphics side of things in these images (of which there are 3!). Instead, this was the first run of converting assets made by artists and rendering them in game. A pretty important test since I was likely not making assets in the most correct way with my impressive 3D modelling skills. There are 3 images here because the artist that made them (Ian Hampton) was unhappy with the angles I was taking the screen captures from. He wasn’t happy with any of these, but I had to get back to work.
These images were taken on 8/15/12.
GoodGraphics07.png, an artist made skull model.
GoodGraphics07_2.png, an artist made skull mode from a different angle.
GoodGraphics07_3.png, an artist made skull model in wireframe.
Around this time, summer school was wrapping up and I got pretty busy prepping for all of my finals and projects due at the end of the semester. Unfortunately, that also meant that there wasn’t a lot of graphics work getting done at the time. The big update here is that these pyramids are a mesh that I built in Max and exported through the FBX conversion pipeline that I wrote for our engine. Each pyramid is 8k polys, and there are a lot of them in the scene.
This image was taken on 8/14/12.
GoodGraphics06.png, terribly high polycount models exported from Max.