About UV mapping

The game I’m now working on is my first real attempt in 3D and other than working on the design and mechanics of the game, in the last couple of weeks I’ve been learning a lot about creating UV maps to apply textures to 3D models. I had little knowledge and negligible experience in the field so I decided to spend some time to master, so to speak, the matter.

In case anyone’s wondering:
UV mapping is the 3D modelling process of projecting a 2D image to a 3D model’s surface for texture mapping. The letters “U” and “V” denote the axes of the 2D texture because “X”, “Y” and “Z” are already used to denote the axes of the 3D object in model space. (from Wikipedia)

On the left an image showing the actual UV map of the 3D model on the right, a low poly spaceship model I’m using in my game, it’s just 56 faces.

Kill that polygon!

Along the journey I ended up improving my 3D modeling ability, especially regarding low poly models, something quite important if you’re trying to create a 3D game aimed to run decently even on older devices running on Android 4+.
I knew it was important but I’m now aware of the fact that lowering the number of polygons in a 3D model as much as possible without ruining the model itself is quite a form of art.
If you are going to create your own models then you have to be careful while making them, for example you might be tempted to use curved surfaces, they’re soo cute… but after all we are talking about making models for apps that will run on devices with small displays like smartphones and most tablets.
On a small screen there will be little difference between a curved shape and an approximated or even a non curved one in many cases but you might end up with a tiny fraction of the polygons and this matters a lot in terms of computation required by the CPU+GPU of the device.
Consider the model in the picture, it’s first design contained many curved shapes and was a few thousand faces (a.k.a. polygons), after redesigning it a few times now it is just 56 faces, more or less 1% of the original model.
Sure, the first one was beautifully rounded, while the second one has rough edges, the point is that you really don’t need that level of detail in a model that probably will just fill up less than 5% of a 5 to 7 inches display.
This is especially true considering that the model will be textured (which is: wrapped up by an image) and material/shaders will be applied so the casual viewer’s eye won’t be aware of the fact that your model is low-poly, it will look beautiful if you applied nice textures and shaders on top of it.

The first part of this journey ends here.
More on this in next article:A BIT OF A PITA.


Leave a Reply

Your email address will not be published. Required fields are marked *