What if 3D model voxels/polygons where limited like pixels?

Question

First, let me clarify that I don't understand much about programming, neither about 3D modelling. And I couldn't think or find a better place to ask this, but I hope I don't make too many mistaken assumptions.

But I'm still quite ignorant about 3D modelling in general, I don't even know the difference between then and polygons, for example.

• Pixels and Vectors:

So, from the little I think I know, pixels are limited amounts of information that tells the monitor how to light a certain number of LED's. So if they are resized, they lose quality. While a Vector is somewhat of a calculation of a certain shape and/or colour (I think), so even if they are resized, they won't lose graphical quality.

• About Voxels/polygons:

So, from what I could understand about Voxels/Polygons is that they are more "closer" to Vectors than to pixels in the sense that they can be re-scaled, twisted and textured in any way you want without losing quality.

• The Question:

So, taking into consideration what a pixel and a Voxel is, it would be possible to make Voxels work more (if not the same) as pixels?

What I mean is that, while Voxels are more or less the calculus of a three-dimensional shape in a space, there could be a way of making this three-dimensional shape be a limited amount of information? Or this is more closely related to Polygons?

My conception is that since these 3D information are not changeable, just like a pixel, you wouldn't need as much processing power to render these things. But you wouldn't be able to do much either, since they wouldn't be able to be interacted with like a voxel.

I don't know if I was clear enough, but let's just say that the idea is to fill a certain space with X number of pixels, but three-dimensional, without the use of 3D model programs.

Like a 3D jpeg image, but without the "filling".

• The (possible) function:

Well, while you wouldn't be able to make day and night cycles, solids, fluids and whatsoever, it could be used in really niche applications where you really don't want to use a lot of processing power. Just like a flash game, so to speak.

Or Voxels are already just like this?

Answer 1

One useful way to create and utilize voxel data is to have the underlying data of an object simply be a list, array, vector, whatever, of numbers, colors, etc. From this voxel data, which is analogous to a 3d version of a 2d image, you can create a renderable mesh made of polygons using some algorithm like Marching Cubes or perhaps like Minecraft: loop over all the voxels, for each one check the adjacent voxels, and for every adjacent clear space add a quad (or two triangles) polygon with the color and/or texture of the corresponding voxel data. Either way, short of using raytracing (which is a simple concept to understand but can be pretty advanced to implement), you will likely store the data as voxels but create a mesh around this data made of polygons to render it.

To directly answer the question, voxels (when rendered as cubes) can behave exactly like pixels and act as a sort of 3d image. The voxel data is not necessarily the same as what gets rendered on the screen, however; think of it as more of a point cloud of numbers, that you have to wrap a mesh of your choice around. There are smooth ways to do this, like Astroneer's terrain, or more blocky, such as MagicaVoxel models and Minecraft terrain.

What you're describing is exactly how voxels already behave, but understand that short of raytracing you will have to use 3d Polygon rendering to display the data.