This post will be about 3D Game Models YAAYY!

So lets start off by first a little bit of theory. Firstly there are 2 areas of modeling. One is modeling for RENDERING, and one is modeling for GAMES. When we model for Rendering, we are modeling at the highest quality possible because our end product will be used for a movie or television show or other graphical application OTHER than a video game, OR for CINEMATICS in a video game. Modeling for rendering takes up A LOT of memory, space, time, etc. To give you an idea, a single frame can take as much as an HOUR to render to DAYS depending on what is being rendered! In these cases, many frames are done simultaneously by many systems so that the rendering can be done fast.

Modeling for video games is VERY different as we are concerned about REAL TIME RENDERING! The amount of triangles your model has can greatly effect a games performance! This being said, it will take a lot of tweaking to get the right balance between game speed and graphic quality (number of polygons) equal so that you get the maximum speed and the maximum graphic quality (number of polygons) possible.

Below is a technical chart showing the balance between these two areas:

As you can see, depending on what kind of game you’re creating, you’ll want to model a certain way.

Now remember, as game creators we need to decide what platform we are programming for and from a business perspective, that will determine where on the chart we want to model for. For example if we are creating PC games, it would be great to program in the blue area at the top level graphic level (And by graphics I mean polygon count, that doesn’t mean you can’t get the best looking models in lower polygons because you can and there are tools to help you do this) but as a business decision this would not be good as the majority of users and households have a mid range system with an on board graphics card. You’d only be marketing to those who have a graphics card, however most end users have on-board or basic graphics processors. This means that the only people to buy your game would be game enthusiasts or graphics artists, which is not a large market vs the mass public who have mid range computer systems with an on-board graphics card.

Thats the PC market and you want to make MORE sales so you can fund your future game projects! If however you are coding for CONSOLES like Playstation or XBOX then higher graphics will be fine as your market is now on XBOX or Playstation users who will all have the same hardware.

Most Indies have their first games on computer as thus this is the area that most indies start creating games for. Hence why the RED area is the best for this which will hit most systems AND get great gameplay and quality at the same time.

If you’re making a 2D game, then you can choose the lower systems as well.

Okay so now that you’ve decided on an area to model for, (In Macrion: The Beginning, we are focusing on the Red area as our first launch will be for PC only) we will now want to decide on model level counts.

In game theory the idea is that the further we move from an object, the smaller it gets, so there is no point in rendering 10,000 triangles when the object is occupying only a few pixels on the physical screen! You might as well just have those 5 pixels lit up with the colors of the object which would save your processing speed which would mean a MUCH faster game!

That being said, in the game modeling world we have a term called LOD or Levels Of Detail! THIS IS VERY IMPORTANT! Without having various levels of detail, a game can slow to a crawl and become unplayable, even if it has the best graphics. Game play is important and without it, a game is not a game!

So now in Macrion we have model polygon (Triangle) targets:

5-10k max polygons: Main game character
1-2k max polygons: NPC (Non-Player Characters)
1k max polygons: Level 1 detail (highest detail for game)
750 polygons: Level 2 detail (mid-high detail for game)
500 polygons: Level 3 detail (mid level detail for game)
250 polygons: Level 4 detail (low-mid level for game)
100 polygons or less: Level 5 (low level for game)

So what this means is that as the object moves away from the camera, we want to display the lower polygon version of the object and as the object comes closer to the camera, we display the higher version. This will help us in making the game move faster than if we were rendering all polygons at the same time.

Here is a flower that will be in Macrion: The Beginning with all 5 levels of detail (As well as the first image as the original model with 6000 polygons approximately) close up (We will also display other models from the game as we make them in future posts so users can see our progress):

As you can see above, the same flower is shown with the various levels there. We use Milkshape 3D to create all our models and animations and in a future post I will go over how to actually create this flower from scratch using reference images.

To give you an idea, the reference images we used to create the flower are below:

Okay so that in a nutshell is modeling for game programming. Modeling for rendering is a different thing all together and this is where 3d artists will differ. The tools used to create models are different based on what you are doing and the model quality will differ greatly based on that.

Hope that helps give you an idea of modeling for game creation.

Until next post!

Nav