The generation of accurate reflective band imagery is complicated by the intrinsic properties of the scene, target, and camera system. Unlike emissive systems, which can be represented in equivalent temperature given some basic assumptions about target characteristics, visible scenes depend highly on the illumination, reflectivity, and orientation of objects in the scene as well as the spectral characteristics of the imaging system. Once an image has been sampled spectrally, much of the information regarding these characteristics is lost. In order to provide reference scene characteristics to the image processing component, the visible image processor in the Night Vision Integrated Performance Model (NV-IPM) utilizes pristine hyper-spectral data cubes. Using these pristine spectral scenes, the model is able to generate accurate representations of a scene for a given camera system. In this paper we discuss the development of the reflective band image simulation component and various methodologies for collecting or simulating the hyperspectral reference scenes.