Pre-rendering

Pre-rendering is the process in which video footage is not rendered in real-time by the hardware that is outputting or playing back the video. Instead, the video is a recording of footage that was previously rendered on different equipment (typically one that is more powerful than the hardware used for playback). Pre-rendered assets (typically movies) may also be outsourced by the developer to an outside production company. Such assets usually have a level of complexity that is too great for the target platform to render in real-time.

The term pre-rendered refers to anything that is not rendered in real-time. This includes content that could have been run in real-time with more effort on the part of the developer (e.g. video that covers a large number of a game's environments without pausing to load, or video of a game in an early state of development that is rendered in slow-motion and then played back at regular speed). This term is generally not used to refer to video captures of real-time rendered graphics despite the fact that video is technically pre-rendered by its nature. The term is also not used to refer to hand drawn assets or photographed assets (these assets not being computer rendered in the first place).

Advantage and disadvantage

The advantage of pre-rendering is the ability to use graphic models that are more complex and computationally intensive than those that can be rendered in real-time, due to the possibility of using multiple computers over extended periods of time to render the end results. For instance, a comparison could be drawn between rail-shooters Maximum Force (which used pre-rendered 3D levels but 2D sprites for enemies) and Virtua Cop (using 3D polygons); Maximum Force was more realistic looking due to the limitations of Virtua Cop's 3D engine, but Virtua Cop has actual depth (able to portray enemies close and far away, along with body-specific hits and multiple hits) compared to the limits of the 2D sprite enemies in Maximum Force.[1]

The disadvantage of pre-rendering, in the case of video game graphics, is a generally lower level of interactivity, if any, with the player. Another negative side of pre-rendered assets is that changes cannot be made during gameplay. A game with pre-rendered backgrounds is forced to use fixed camera angles, and a game with pre-rendered video generally cannot reflect any changes the game's characters might have undergone during gameplay (such as wounds or customized clothing) without having an alternate version of the video stored. This is generally not feasible due to the large amount of space required to store pre-rendered assets of high quality. However, in some advanced implementations, such as in Final Fantasy VIII, real-time assets were composited with pre-rendered video, allowing dynamic backgrounds and changing camera angles. Another problem is that a game with pre-rendered lighting cannot easily change the state of the lighting in a convincing manner.

As the technology continued to advance in the mid-2000s, video game graphics were able to achieve the photorealism that was previously limited to pre-rendering, as seen in the growth of Machinima.

Other Languages
français: Pré-rendu