Final Fantasy VII released in 1997 with a technique that confused players for years: the backgrounds looked stunning but the characters looked blocky. Square rendered the environments on high-end workstations, baked them into static images, then overlaid real-time low-polygon character models on top. Cloud Strife's field model had roughly 900 polygons. The Midgar slums behind him were a painted photograph of a 3D render.
This was a deliberate production decision, not a hardware limitation they stumbled into. The PlayStation could not render those environments in real time. So the art team pre-cooked them at full quality and locked the camera angles to match. Every walkable path, every shadow, every light source was calculated once and embedded into the background image.
Freelancers doing environment work for indie studios sometimes get asked to replicate this look. The actual workflow involves rendering your scene in Blender or Maya at high resolution, exporting a flat image, then building collision and character layers separately in the game engine. It is more work than it sounds because your lighting decisions at render time are permanent.
Understanding why Square made this choice helps you explain the trade-offs to a client. Prerendered backgrounds save real-time performance but remove camera flexibility entirely. That conversation, grounded in a specific 1997 example, positions you as someone who thinks through production consequences rather than just executing tasks.