Introduction to the ARS project.

Welcome to my blog, on this page you'll find development updates for this project being developed as my bachelor's final. While the goal for now is getting a working prototype for December, I do intend to keep this project up to date if it performs well.

Now to the meat and gravy of it... "what is this ARS thing?"

It stands for Agnostic Rigging System.

The goal is to develop a functional rig generator, that will automate quite a few things, but here are the big lines:
 

1. Game Engine Integration: This means that most of the simulations and features will be generated in the game engine rather than have them calculated at the animation software level. Things such as arms roll-bones, muscle deformations or simulated physics will be produced only in the game engine and tweakable there, keeping the 3d package clean of it all to improve performance and mobility of the file format.
2. Software Agnostic Approach: At this point, I want to be able to generate this system in either Maya or 3ds Max and have anything generated in one be transferable to the other one seamlessly. Nowadays in productions, we seem to be forced to decide which tool will be used by all simply because of a lack of communications between those interfaces. I would like to bridge that gap with a proof of concept. It could be to keep costs low (imagine sending a complex rig down to Blender for instance, reducing the amount of licences).
3. A form of rig-building system: Just trying to compete with the openly available tools out there. :D

There are many challenges regarding this sort of system... for instance, a pipeline module in it's own that would need to be attached to any sort of existing production pipe. The current sketch of such a pipeline would be this one:

1. Master Scene: this one is rather easy... this is where you start building the parts of the rig. whilst this system is relatively simple, Unreal 4 (current project holder) will need to to have a complete rig done in order to accomodate all the bones/joints, simulation informations and so on. This scene would be creatable either in Maya or 3ds Max in an ideal world, and regenerate itself towards the opposing software when you generate it, translating it's functions from one side to the other depending on the supported features.
2. Anim Scene(s): those scenes are generated at the export of the Master Scene. This scene includes a lightened version of the rig, as in, it contains all the necessary controlers, but none of the calculations behind the scenes. It will remove all the unnecessary calculations or bones, apply a rather simple skin to it all and have an approach on functionality. Hopefully I'll have time to make a module to save animations in a format that all applications can easily translate and re-apply as needed. It should be noted that this scene would contain a refferenced rig, so that any updates to the Master Scene will always propagate down to the animations, and the animation save system would reapply locally with minor tweaks to be done.
3. Unreal Integration: This part would include macros to generate all the desired features, perhaps a pluggin to make all the anim graph settings to setup automated rollbones, muscle simulations, collisions and ragdoll switches for dynamic animations... so on. Again this being a working proof of concept first, this will most likely start appearing later in the process. Mobility between apps is the goal for now.

All in all, like all good rigs, the most important aspect is the same as it normally would be: Ease of Use and as many automated systems for added realism.

I'm confident I can make this happen!