Pre-visualization, commonly shortened to previz, is the process of creating rough visual representations of planned shots, sequences, or entire productions before principal photography or full production begins. Previz allows directors, cinematographers, and visual effects supervisors to explore, test, and communicate complex sequences in a low-cost, easily revisable form before committing to the expense of actual production.

Previz takes many forms depending on the production's scale and needs. Animatics use timed, edited storyboard panels or simple animations cut to audio to test pacing and editorial flow. 3D previz uses rough computer-generated environments and character stand-ins to explore camera positions, movements, and staging for complex action sequences or visual effects shots. Techviz, a more detailed variant, calculates precise technical requirements like camera positions, lens choices, and equipment placement for effects sequences that demand exact physical coordination. On larger productions, dedicated previz teams work in parallel with pre-production planning, delivering visual sequences that directors use to plan shoots and communicate intent to department heads. The value of previz lies in its reversibility - bad ideas are discovered and fixed before they consume production resources.

AI generation has become a powerful previz tool, enabling rapid visual development of sequences at a fraction of the cost of traditional 3D previz. Generating rough versions of planned shots to test compositions, lighting directions, and visual approaches before committing to full-quality production - using generated clips as a visual plan that guides the final generation workflow - applies the core discipline of previz thinking to AI-native production.