Video Game Design/Programming/Framework/Procedural Content Generation
Procedural Content Generation
[edit | edit source]The concept is defined as the process of procedurally generating any type of 2D/3D geometry, textures, scripts or sound, even using a randomization seed. To date this technique is primarily used to create environments, simple worlds or structures. This approach can also be applied to many other content, using strict definitions it can be used in cooperative work with AI (dynamic environments, object and creatures), but it is still difficult to be as reliable as realistic as traditional animation, even if it can provide greater variability and in replicating human voice it is still at its infancy.
Unless there is a benefits for interactivity or to enable randomization, procedural content generation should be avoided as it is expensive in terms of CPU cycles and generally complex to implement, exceptions are made to repetitive but simple patterns, even if interaction is not possible, for instance in the simulation of water, the effect of rain or even the display of clouds, smoke, fire or explosions. A less important factor that may prompt the use of procedural content generation are space (size) restrictions in the finished product, due to the media to be used or the size of the download.