Ray tracing is one of the photorealistic rendering method. Ray tracing process takes a long time because of the calculation of lighting effects and optical effects caused by the material of the object. The length of rendering time increases with the increasing complexity of the shape of the object, such as a mesh object. To minimize processing time, generally done by reducing the complexity of the mesh, which reduces the number of face that forms a mesh object. How to reduce the number of face is to combine adjacent face to face. The downside of this method is that the object surface resulting display will look grainy/triangular - facet (checkered). The proposed method in this study aims to refine the visualization of objects that complexity is not high. Triangular surface mesh - terms appear due to the calculation of the same color for each face that form a mesh. If the face has a color gradation, the mesh will look smooth surface. To produce a color gradation on a broad face, normal calculation was used, ie in proportion adopting b - splines curve calculations. Testing is done by comparing the results of the use of normal B - splines to the simple method of subdivision (a method of face vast divide into several smaller face , so that the finer the mesh surface) with ray tracing rendering method. From the experimental results, it can be concluded that the results of ray tracing using b - splines normal when compared with the subdivision, the results are not too different. From the aspect of long rendering process, the more the number of face of a mesh, it also increase the effectiveness of rendering time process. However, the increase is not proportional to the increase in the number of mesh