Animation Skills - Final Animation
Final Project
We were asked to produce an animation using all our previous knowledge to be modelled and animated in Maya using our previous exploratory vehicle.
To ensure that the animation continuity was correct, we started by creating a storyboard. The storyboard simply concatenates shots for the timeline.
With a completed storyboard, the next stage was to obtain a rigged character to animate. We were given a basic pre-rigged character to use, although I felt this was not suitable for my standard of animation and thus searched for a more realistic character.
I started by animating a walk and chose a 24 frames walk cycle and as an animation runs at 24 fps, this will equate to approximately one walk cycle every second. This speed is roughly typical of the speed most people walk and thus looks quite realistic.
The next section of animation was interacting with my exploratory vehicle where the character climbed a ladder inset into the hull. For this I had to create the ladder and also a hatch at the top which was able to be animated up and down. The climbing animation was interesting because it involved moving the upper arm backwards and the forearm forwards at the same time to raise the hand without moving it too far forwards into the hull. There was a small problem with gimball lock on the rigging which involved manually adjusting several series of frames.
After animating the climb and the character standing up on the hatch ready to descend into the vehicle, I played this back and noticed that the animation was very slow for the first 21 frames and needed to compress all 21 frames into 7 frames. To do this I used the dope sheet which allowed me to rearrange key frames interactively. Once this was done, I animated the hatch and the character together going down and then the hatch coming back up without the character.
The final sequence of the animation was to animate the 3 propellers as the character supposedly starts the engines. This was in actual fact tricky to achieve a realistic effect as any engine would start slowly and gradually increase speed. Thinking about algorithms and number sequences, one in particular caught my eye and that was the Fibonacci sequence (1, 1, 2, 3, 5, 8, 13, 21, 34, 55, etc), this seemed to match the requirements and I used this to increase the angle of rotation of the impellers at a constant frame rate. Thus starting from frame 0 at 0 degrees, then frame 5 at 1 degree, frame 10 at 2 degrees, 15 at 3 degrees, 20 at 5 degrees, 25 at 8 degrees, etc. This proved a little slow to start with and thus the sequence was started at 3 degrees and continued until the speed was deemed to be fast enough.
Following this, I created skydome light, added textures to the walkway and vehicle before creating different cameras to render from. I initially tried batch rendering but the student version of Maya does not include an Arnold licence and thus my render had watermarks all over the screen which wasted several hours trying to find a way to render without the watermark eventually settling upon the render sequence command. Initial timings of frame creation suggested around 7 hours of processing time on my computer to render around 400 frames and thus a decision was made to split rendering into several sections and use several computers reducing the time taken by about 50%. I am writing this blog as rendering continues and any typographic errors will probably be due to my typing many characters without being able to see anything until the document window gains processor and graphics time.
I feel I have done reasonably well on the animation. My chief annoyance was that of the vehicle which was constructed in a monocoque fashion to achieve hull pressurisation for submergence in liquid. Due to this fact, it was difficult to provide an interior easily and quickly or indeed an opening door and thus I had to improvise and use a hatch. Strictly speaking the hatch should be flush with the monocoque and have a bevelled edge to keep pressurisation. The climbing sequence although reasonably good, could be better by using many more rigging movements to achieve a realistic climb and there are portions of the climb I could do much better given the time. Looking at the walk animation, I find my character has anti gravity boots as the walk is conducted some 5 mm above the surface of the planet which was difficult to see until final rendering and I decided not to do another 4 hours rendering just to correct this and instead have decided that anti gravity boots are an adequate solution in this case.
UPDATE: - Animation has finally finished on the second computer after 9 hours of processing at 100% cpu on a Core i7 with 4 threads running at 3.6 GHZ and running an Nvidia GTX 970 graphics card.

Comments
Post a Comment