Day Three: Even More Rocks!

So today we went to reshoot the entire movie, but with a fast SD card this time! There were some problems during the shoot today, like the sun randomly getting blocked by clouds, and some depth of field problems with the matchmove shots. For those shots, I increased the shutter speed and had to open the aperture wider to compensate, however this resulted in the markers being blurred out of focus. And it starting raining when we were almost done. And so we'll have to reshoot, again. Fun times. I was the director of photography for these two days but Praveen would be taking over tomorrow.

After giving some feedback to improve the shots, Mr Ron introduced us to camera tracking with Nuke. At last! No more Matchmover!

Some progress on the rock generator. I've added a crackle multiplied with turbulence noise for the little details. I was also experimenting with ways to have a flat top for the rocks. At first I tried grouping the top points together and exclude them from the subdivide and VOP SOP. However, I had encountered some strange stretching problems and it wasn't procedural. I then tried using a cookie, which kind of gave me what I wanted, but Mr Ron suggested a better, which was to use a clip , then group the unshared edges together and applying a polycap. This gave the stretching problem too, and we discovered that it was due to the face not having a normal. A group geometry node was used to procedurally select the last face, and add a normal to it in the Y direction. This gave me exactly what I needed.

HEHE