Computer-generated imagery, commonly abbreviated as CGI, is imagery created or substantially generated using computer software rather than being captured entirely through conventional photography or physical artistic processes. CGI can be two dimensional or three dimensional and can range from simple graphic elements to highly realistic digital environments, characters, objects, simulations and entire animated films. Three dimensional CGI commonly involves modelling objects or characters, assigning materials and textures, establishing lighting, positioning cameras and rendering the scene into images.
CGI is widely used in films, television, games, advertising, architecture, engineering, scientific visualisation and virtual reality. The term does not necessarily mean that every pixel of an image was generated by a computer. A CGI element can be combined with photographed material through compositing, and computer generated imagery can also be deliberately stylised rather than photorealistic. CGI is therefore a broad production category covering computer created visual content rather than a single visual style.
Three dimensional CGI generally begins with modelling, the process of constructing a digital representation of an object or character. Models can be created through polygonal modelling, digital sculpting, procedural techniques, parametric construction or combinations of methods. Polygonal models consist of vertices, edges and faces that define a surface, while sculpting systems allow artists to shape forms using digital tools similar in concept to physical sculpting. The complexity of a model depends on its intended use.
A close up film character may require extremely detailed geometry, while a distant background object may need far less. Topology refers to the organisation of the model's geometric structure and is particularly important for objects that deform during animation. Efficient topology can support clean deformation, predictable subdivision and manageable computational requirements. Modelling therefore establishes the geometric foundation on which later CGI processes depend.
Materials and textures determine how CGI surfaces appear when illuminated. A texture can contain information about colour, roughness, normal variation, displacement or other surface properties. Modern physically based rendering systems commonly describe materials through parameters such as base colour, metallic response, roughness and specular behaviour, although exact terminology and implementation differ between renderers. A material is not simply a painted colour because surfaces interact with light according to their physical characteristics.
Metal reflects light differently from skin, glass behaves differently from fabric, and rough stone scatters and reflects light differently from polished plastic. Texture artists therefore create maps or procedural patterns that provide surface variation. Accurate material representation is especially important for photorealistic CGI because viewers are highly sensitive to familiar physical properties.
Rigging and animation allow CGI objects and characters to move. A character rig provides a system of joints, controls and deformation mechanisms that an animator can manipulate without directly editing every vertex. Skinning associates the character's surface with an underlying skeleton so that joint movement produces corresponding deformation. Additional systems can control muscles, facial expressions, cloth, hair and other secondary elements. Animators then use keyframes, motion capture, procedural systems or combinations of techniques to create movement.
Mechanical objects may use constraints and precisely defined transformations, while organic characters require consideration of anatomy, balance and deformation. Animation is therefore a separate stage from modelling even though the two are closely connected. A beautifully modelled character may still require substantial rigging and animation work before it can function convincingly in a moving scene.
Lighting and rendering convert a digital scene into a finished image. Virtual lights can be positioned and configured to simulate sources such as sunlight, lamps, studio fixtures or environmental illumination. Rendering calculates how light interacts with geometry and materials and produces pixel data for the final image. Modern renderers may use physically based calculations such as ray tracing, path tracing or other numerical methods to estimate direct and indirect illumination, reflections, refractions, shadows and global lighting effects.
Rendering can be computationally expensive, particularly for high resolution images containing complex geometry, volumetric effects, hair or multiple light interactions. Production pipelines therefore balance visual quality against render time and hardware resources. The rendered image can then be passed into compositing, where it is combined with other elements and adjusted for the final shot.
CGI can be photorealistic or deliberately stylised. Photorealistic CGI attempts to reproduce the appearance of photographed reality through accurate modelling, materials, lighting, camera behaviour and rendering. Stylised CGI may instead use simplified geometry, exaggerated proportions, non realistic colours, illustrative shading or deliberately artificial movement. Animated films frequently employ stylised techniques because the goal is expressive design rather than photographic duplication. The distinction between CGI and realism is therefore important.
CGI is a method of generating imagery, not a synonym for realism. Similarly, not every digital image is CGI. A photograph that has been colour corrected digitally remains fundamentally photographic, while a three dimensional rendered object is computer generated even if it is designed to look like a photograph. Hybrid imagery can contain both categories simultaneously.
CGI is used in many applications beyond entertainment. Architects use computer generated visualisations to communicate proposed buildings and interiors. Engineers use digital models and visual simulations to analyse designs. Scientists use visualisation to represent data, structures and processes that may be invisible or difficult to photograph directly. Medical imaging can transform measured data into three dimensional representations, although such systems involve specialised scientific imaging rather than conventional entertainment CGI.
Games use real time CGI to generate interactive worlds, characters and effects. Virtual and augmented reality rely heavily on computer generated environments and objects. Advertising uses CGI to create products, environments and camera movements that would be expensive or impossible to photograph. These applications demonstrate that CGI is fundamentally a method of visual representation and computation rather than a medium restricted to films.
CGI has transformed visual production by separating many visual elements from the physical limitations of cameras and sets. Artists can construct environments that do not exist physically, change object proportions, control lighting after the scene has been built and create creatures or events without requiring physical equivalents. However, CGI also has limitations involving cost, computing resources, production time, technical complexity and the difficulty of achieving convincing realism.
Photorealistic human faces can be particularly challenging because viewers are highly sensitive to subtle inaccuracies in eyes, skin, movement and facial expression. CGI is therefore not automatically superior to practical techniques. Modern productions often combine practical photography, miniatures, physical effects, digital models, simulations and compositing. The most useful understanding of CGI is consequently as a flexible visual production technology whose effectiveness depends on appropriate modelling, materials, animation, lighting, rendering and integration with the surrounding image.
