This paper presents an extension of Catmull's algorithm for rendering images of bivariate surface patches, focusing on texture simulation and lighting models. The algorithm uses a coordinate system defined by the parametrization of a patch to map patterns onto surfaces. The intensity of each picture element is computed as a weighted average of regions of the pattern definition function, with the weighting function determined by digital signal processing theory. The algorithm allows accurate computation of surface normals, enabling the simulation of mirror reflections. The amount of light coming from a given direction is modeled similarly to texture mapping and added to the intensity from texture mapping. Examples of synthesized images using these techniques are included. The paper also discusses texture mapping, where the algorithm's ability to simulate textured surfaces is utilized. The bivariate patch is a mapping of the unit square in parameter space, and the coordinates of the square can be used as a curvilinear coordinate system for the patch. The subdivision process tracks the parameter limits of each patch fragment, yielding parameter values at each picture element. These values are used to map patterns onto the surface. The texture pattern is sampled and the intensity is averaged proportionally to the amount of the picture element covered by the patch fragment. This method reduces aliasing, a common issue in texture mapping, by applying a controlled blur to the pattern. The paper also addresses reflection in curved surfaces, discussing lighting models that incorporate terms to produce highlights on surfaces. The model uses a virtual light source to simulate highlights and accounts for the glossiness of the surface. The simulation of reflections in curved surfaces requires accurate surface properties and normal vectors. The subdivision algorithm provides accurate surface normals, enabling the simulation of mirror reflections. The direction of the reflected ray is determined using the surface normal and viewing position, and the reflected light intensity is read from an environment map. The techniques for texture and reflection can be combined to produce images of objects with patterned shiny surfaces. The combination of highlighting and texture mapping scales only the component from the real light source, modeling the highlight as specularly reflected. The techniques are applied to generate images of a highly glazed patterned teapot. The paper concludes that refining and extending Catmull's subdivision algorithm allows for the generation of images with a higher degree of naturalness. These improvements result in better techniques for generating patterns and textures and new capabilities for simulating reflections.This paper presents an extension of Catmull's algorithm for rendering images of bivariate surface patches, focusing on texture simulation and lighting models. The algorithm uses a coordinate system defined by the parametrization of a patch to map patterns onto surfaces. The intensity of each picture element is computed as a weighted average of regions of the pattern definition function, with the weighting function determined by digital signal processing theory. The algorithm allows accurate computation of surface normals, enabling the simulation of mirror reflections. The amount of light coming from a given direction is modeled similarly to texture mapping and added to the intensity from texture mapping. Examples of synthesized images using these techniques are included. The paper also discusses texture mapping, where the algorithm's ability to simulate textured surfaces is utilized. The bivariate patch is a mapping of the unit square in parameter space, and the coordinates of the square can be used as a curvilinear coordinate system for the patch. The subdivision process tracks the parameter limits of each patch fragment, yielding parameter values at each picture element. These values are used to map patterns onto the surface. The texture pattern is sampled and the intensity is averaged proportionally to the amount of the picture element covered by the patch fragment. This method reduces aliasing, a common issue in texture mapping, by applying a controlled blur to the pattern. The paper also addresses reflection in curved surfaces, discussing lighting models that incorporate terms to produce highlights on surfaces. The model uses a virtual light source to simulate highlights and accounts for the glossiness of the surface. The simulation of reflections in curved surfaces requires accurate surface properties and normal vectors. The subdivision algorithm provides accurate surface normals, enabling the simulation of mirror reflections. The direction of the reflected ray is determined using the surface normal and viewing position, and the reflected light intensity is read from an environment map. The techniques for texture and reflection can be combined to produce images of objects with patterned shiny surfaces. The combination of highlighting and texture mapping scales only the component from the real light source, modeling the highlight as specularly reflected. The techniques are applied to generate images of a highly glazed patterned teapot. The paper concludes that refining and extending Catmull's subdivision algorithm allows for the generation of images with a higher degree of naturalness. These improvements result in better techniques for generating patterns and textures and new capabilities for simulating reflections.