Thursday, 7 November 2013

Creating Cyclic Animations for Video Games.

So I wasn't able to complete my 11 second club submission. Very depressing.

My terrible time management skills are mainly to blame for this. I really need to establish some sort of timetable for working on each module, because I'm finding that by meandering between modules on a whim I'm often left spending too much time on certain areas.

But anyway, back to animation.

As I think I mentioned before this coursework is focused on creating cyclic animations for video games, and as such the animations should pertain to ones you would expect to see in games, and should be realised through an appropriate video game character rig.

Now, the other week I had my first mo-cap session. Mo-cap is something that has interested me since the VFX trip to London I went on back in 2012. Then recently, I attended a talk by Jonathan Cooper, the lead animator on Assassin's Creed 3.

The talk was titled "Animating the 3rd Assassin" and in it he detailed his mo-cap & animation process for the Assassin. The quality of animation that can be achieved blew my mind. Johnathon explained the process, which involves skilled actors and state of the art suites, and how he directs the mo-cap process, applies it to a mesh, and then tweaks and refines the data it to create one awesomely animated character.

Towards the end of the motion capture session I was allowed to direct a dying animation - which I really enjoyed. And the almost-immediate feedback you receive with mo-cap really lets you iterate like crazy.

I'd like to book out the motion capture suite sometime soon, but before I do that, and this was heavily pointed out, I need to meticulously plan everything. Just as a producer for a game of film considers time and budget, it's great to get into this mode of thinking and practicing professionally.

In class we were shown Johnathon's showreel, which included a walk, a sneak, a run, and a breathtakingly impressive quasi-free-running sequence.

This array of animation really makes sense for the Assassin.

That's what I feel made the showreel really pop. The animations are so strongly rooted in the context and personality of the character.

So, to begin an interesting workflow, and also allow me to use my time as cleverly as possible, I'm going to be tying in this animation module with my Low Poly Modelling class.

I've never actually done this before, although it's often suggested. So I guess I'm quite excited about, because it means I'll be more focused on mutual assets, which I reckon can only be a good thing. Quality over quantity and all that.

So yeah, the reasoning for tying these modules.

Well, for Ken's module I'm interested in the translation of both the transition of stylised concept art to 3D, and also the way in which a characters' adult character model compares to the child counterpart. I think this leaves a nice scope to take the 3D model of the child and the adult and use cyclic animation to convey their individual personalities. It would also provide a good chance to provide context to the characters animations. I think using the characters' back-stories, which would be devised in LPM through the concept art, will provide interesting situations which can in turn be animated for this module. Hopefully I'll get around to re-wording this as I feel I might be rambling or perhaps unclear.

I guess for rigging I'll need to look into using Mixamo or something similar to get the rig. If not, i'll just use the essence of the characters, and a downloaded rig. Naturally I'd prefer to use the actual models.

That's all for now, folks.

No comments:

Post a Comment