What options do I have for a 3D game without graphics but yet maintaining physics for later ?

I don't believe I expressed well the title for my question so if you believe you can change it for a more suitable title please do so or drop me a comment.

Currently I am starting a 3D game with C# but I don't have a designer yet nor any images or anything or any money to invest on it.

• I would like to develop it using some sort of placeholder or whatever to fill the graphics so I can make at least the basics of it and later simply switch for the real thing. What tools should I take into consideration for this job that are non-commercial and/or that won't limit me once I have the game ready to release in terms of paying the 3rd parties for extra resource or whatever?

• Should the place holders be actual sample images or can I use for example a ball to represent a body (what will be the impact of it later when I switch to the actual body it self) ?

From game development Prix

• The easiest placeholder graphics are those that you just whip up in code. For a 3D game, write something code to generate a plane, a box, a sphere, a cylinder. Bam, now you have enough stuff to make a game!

  Prix : Thanks for the reply, that is what i was thinking of. Like using a box or a sphere to represent a body or something alike but what I am really concerned here is about the later conversion from whatever I used as placeholder to the actual graphics. For example using a box to represent a person's body what will I have to consider later when I convert it to the actual body, etc.

  The Communist Duck : I don't think you would have to consider anything, as long as you treat the box as a body rather than a box.

  dash-tom-bang : @Prix - later you just put in different visuals and update the collision model if it becomes important for some reason. Don't worry about later, though; software can be changed.

  Prix : @dash i got the point i just didnt wanted to have to work it all over once it is done...

  dash-tom-bang : @Prix the point is that you don't have to work it all over once it's done. If the initial solution looks fine, just keep it. You don't need to draw the boxes, you can just use them for the physics. You don't need to rewrite anything if your eventual character uses the same code as the renderer you write for these primitives.

  Prix : @dash once again thx.

  Olie : Quick add to this: if your whatever-game-object has moving parts, go ahead & toss a box/sphere/whatever in there for that, too, so you can code-up the movement. So, for example, you might make a person with a sphere as head (that rotates to face another direction), 2 box-arms that go up & down, and 2 box legs that move back & forth, or whatever. Point being, if "make walk motion" or "nod head" are parts of your game, it's good to be able to see representations of that. Later, you can replace all your models with "nice" ones, and get the desired effect.
  From dash-tom-bang