Week 10

This week Sav had a finished animation of the pistons moving inside the engine. At first we were all very impressed with how the animation had rendered. After much closer inspection we started to realise that the last piston in the sequence was jamming. It didn't take long for Sav to realise the problem and we came up with a solution. The reason for the piston jamming was down to the fact that it was the only piston in the sequence that had not been animated to loop once finished. In essence the problem was simple to solve but because of the large amount of frames per second we knew it was going to take a long time to complete.
We overcome the large rendering time by splitting the animation into five sections and rendered them accross five computers. This still took one - two hours per section to render.

The biggest factors effecting the render time were our choice of lighting, the use of reflective materials and the actual renderer we used. We used Mental ray lighting and used the Mental ray renderer to produce uncrompressed 'avi' files . We chose to render the scene as an uncompressed file so as not to lose any quality when importing the video sections into the video editing software.

We also found that when the scene was rendered all of the materials which I had applied to the engine previously had been white washed. The material properties still remained but there was no colour other than white. This down to the settings that had been applied during the lighting process. Although this happened by mistake, we all liked the simplicity of the colour scheme. With everything rendered white we found that textures did not take away from what we actually wanted to show- the pistons.

At this point in the production process it became apparent to us that the more we added to the animation the more complicated and time consuming it became. We calculated that if we used the environment that the render time would take around thirty hours. This wouldn't be possible on the timescale we were working on. The extra modelling would also mean that the file size of the animation would double, meaning it would take up a lot of space on the clients computer.

Further cuts from the animation could include the Spitfire model. We feel that taking these aspects out of the animation will not damage the message we are trying to portray. Without these aspects we can allow for a higher error of margin. If we need to repeatedly render the scene it will take suggnificantly less time than we originally anticipated.

As it stands the animation consists of the engine rotating with one of the piston blocks as a wireframe. Inside the wire frame the pistons have been animated using bones to simulate the linear motion.