How does making a 3D model on my computer put so much strain on it, but rendering it in real time alongside a hundred copies of it in a video game work perfectly fine? Moreover, how does it take so long to render something that was already being shown and worked on?

Question

Hope this doesn’t count as a ‘loaded question’.

In: Technology

Answer 1

I’d assume it’s because you’re making the computer work harder and the machine and it’s drivers inside the computer all work at once. Where as an image is just that, an image so a lot less strain.

Answer 2

There are two factors at play here.

First, and most importantly, model editing software is designed to make editing the model easy, and not to render hundreds of things at once. As such it may use more complex/slower data-structures and algorithms to store and display the model, if it helps with the editing.

The second thing is that games often use lower quality assets (for performance reasons). It is quite common to edit a model in super high quality (simply because it’s easier that way) and then create a much faster version to use in game. The in-game version could for example use normal maps instead of polygons for small surface details, compressed textures and coordinates as well as baked AO or lighting.

Answer 3

Your computer is working hard to figure out where all of the little triangles should actually be. It needs to make sure that they don’t overlap, nothing is inside of the solid and so on. It’s not spending much time actually rendering it.

By comparison video games have already precomputed all of the stuff that matters, and some of it doesn’t even matter (like having points or vertexes inside solids). All it needs to do is actually draw the triangles that have been precomputed. Also games tend to use a lot fewer triangles than you think with textures to hide this.

Once your model actually renders, try rotating it on screen. That should be pretty snappy, right? That’s analogous to what video games are doing.