I've been studying Blender, with the expectation that, somehow, I will get the stuff I create in it into a game. How exactly does Blender and others fit into the development of most games?
I've found that you can load meshes with Ogre, but what about making it move? Say I have a figure, how do I make it walk in the game? Assuming I have the walk animation done.
Animation is just a model that looks different at different times. Exactly how this is achieved and how it is stored depends on what animation technique used and on file-formats and so on. Examples of different animation techniques are keyframe animation and skeleton animation.
I have no experience with ogre but a start would be to search for ogre and keyframe or skeleton animation. That should at least give you a start.
I can find how to do it, that's easy. (Google)
I'm interested more in the theory of how it is used. Are practically all moving(moving, as in, changing shape in some way) objects in games done from animations, or are some done by calling the renderer to manipulate the mesh of the original, static object?
Since you have no experience with Ogre, what do you suggest?
Last edited by User Name:; 03-13-2011 at 06:30 PM.
O_o
Dude, manipulation the mesh would be animation. Look around at all those old "stop motion" films. It is mostly the same thing.
Anyway, every model format I've worked with treats animation a little differently. (This is what Shakti said.) You are best served by poking around mostly standardized formats to get an idea of how the animations are stored within model data.
Now then, I don't know what importers "Ogre 3D" has out of the box, but "Blender" exports a powerful Python API so a crazy number of exporters exist. (The changing API means that some may no longer work, but could always be used a base to build a new version.) If you find a model format you understand, odds are an exporter exists so you can play around with it in code so you can really get into it.
Soma
All movements in games are animations. The only thing I can possibly think of that would modify a model based on runtime input (ie, not knowing the end-result when starting the program, with say walking you know exactly how the model should look at all times when walking) would be if you have partially destroyable objects (buildings you can blow a hole through wherever you wish for instance) or deformable terrain, and even here some trick is probably utilized to make it computationally easy (relatively).
Games of today pretty much dont care how much HD space they take up, it is cheap and can be "wasted". Every single millisecond when running the program counts though, so if something can be saved to HD to make the game run a little faster it will be done.
