In this class we learned more about deformables. The paper/video that I thought was interesting was the Pose-Space Subspace Dynamics Rigging Characters. I thought it was interesting and weird that the resting position they proposed was a hunched over position. But I guess it’s easier for. But also decided to separate different regions of their character models to work on them independently. I got a better understanding of what implicit euler was and how it was used in the Projectile Dynamics: Fusing Constraint Projections for Fast Simulation paper. They would look at the given parameters and then from there go on to the next time step.
There were a couple of papers that I thought were interesting in this class. The top one that I liked was the Fluid Control using the Adjoint Method paper. I thought it was interesting that the methods in this paper were also used in the movie Hellboy 2. That in the Target-driven Animation paper they compared the 2D and 3D fluid simulations and looking at the different computation times. It made sense to me that you had to make the fluid into 2D and then work your way up to a 3D fluid. I’m currently trying to teach myself different kinds of procedural methods through Houdini and other software. I thought that it was interesting that water can also be procedural, and that it is the oldest way of simulating water. (who would've thought!) I definitely want to look more into this as well as the PhysX accelerator that we talked in class that appeared in Siggraph 2007.
Today we had different people come up and present their Mini project 2 projects. I thought most of them would be similar but the majority of the presentations and approaches to cloth simulations were different. One of my favorites was Nicky’s presentation that touched on cloth simulation to music. The frequency of the music correlated with the color of the cloth and could also introduce ripping into it. I was having trouble with ripping and implementing certain cloths and will definitely follow up with Nicky as well as Dun Dun who looked into these topics as well as making it look
Today was our first introduction to fluids. I’ve been waiting on this topic since the beginning of the semester. This along with Rendering are my top animation topics that I wanted to learn about. I appreciated that we took it from a very low level and build or way up on fluids. Certain factors that go into calculating and simulating a fluid is velocity, fluid density, pressure, acceleration due to gravity or animation, and dynamic viscosity. I thought it was interesting that the momentum equation was just a specialized f=ma. It can be used to help model fluids with a bunch of particles such as blobs of water. I also thought it was interesting that we first look at fluids in this case water as blobs and individual particles that are circle and then calculate the entirety of that fluid. That fluids does not only apply to liquids but gases as well.
We went over individual papers that Dr. Pollard suggest for the ideas we presented in class. For my particular project we talked about a lift paper called Realistic Modeling of Bird Flight Animations. This was interesting how they calculated the lift of a bird’s wing in an animation. I’m just trying to figure out how this can be applied to a paper plane. Some part can be applied in terms of flying a little bit but I might have to watch more simulations on paper plane movement to see what parts of the paper I can apply to my project. We talked about Disney’s Tangeled in the paper Simulating Rapunzel’s Hair in Disney’s Tangled, which was pretty interesting to me. It kind of reminded me of some of the challenges Pixar’s Brave movie with the curly hair of the main character.
Everyone presented their pitches for final projects. Some people had multiple ideas and others had a more clear idea of their project. I had a simulation idea that I wanted to try out with a paper plane but I wonder who doable it was with the about of implementation and time it would take. Plus I found out that my idea doesn’t seem to have a lot of papers to back it up. I’m thinking that I might also look into the Boids algorithm for this project as well.
We continued to talk about different kinds of cloth simulations in class. The paper that I liked in this class was the Perceptual Control Space for Garment Simulation paper. One of the CMU faculty Jessica Hodgins also worked on this as well. They tested out different kinds of density on a cloth and put it to human animation. We were able to see how well the garment flowed when a human animation was walking in it. I also like how they took into account of how soft and wrinkled that cloth was.