CG Integration – 6 – Mesh Change

So after some critique from the lovely people at The Mill, it seemed like the mesh and effect created a weird feeling, some even saying “like dog poop”. It’s unfortunate because once it is seen, it can’t be unseen! This drove us to really focus on the material and the mesh being deformed, what makes sense and how would it react? We brain stormed for quite some time and after a bit of collaboration with some friends, we determined a paint bucket would be a great object, it would fit in well, its metal so we could do a cool metal material on it that melts, and it wouldn’t take a lot of time to put together at this last minute. So we made one.

Here’es the mesh before and after the simulation. The simplified mesh on the left had to be quite subdivided to make sure that UV coordinates were properly transferred to the particles before they were sent to sim (see below). Other than that, this bucket looks nice in the scene and was a breeze to model and UV, making this process much more quick and efficient.

The UVs on the left are from the original model before scattering points and simming while the image on the left shows how the UVs look like after meshing the particles and transferring the UV coordinates back to the mesh. There is obviously some loss in quality, however considering the texture itself (how noisy it is) this works perfectly for us. I had to keep the resolution of the mesh quite high for two reasons, so that the UVs would as accurately transfer as possible and so that the rounded shape of the paint can would be kept as accurately as possible. I saw quite a bit of banding across the surface of the can. I could probably scatter the points more randomly than now (which is just an organized grid scatter) but it wasn’t terribly high on my “to do” list, so we will save that for the future.

The material looks really cool, I have to do some more development on the “melted” material, as its a bit too stagnant right now. The sim also looks good, however it definitely needs to be tweaked a bit to solve some problems with the particle separation deletion code I implemented last post.