Recent computing power and network bandwidth becomes sufficient to realize complex 3D virtual environments between distant users. The public enjoys surfing on many complex multimedia web pages via standard personal computers and broadband networks. The creation of realistic 3D worlds, however, is limited to only skilled designers' responsibilities. In this paper, we propose a simple yet powerful 3D authoring method enabling wide range of users to easily create successful 3D models. The proposed method starts a creation process from modifying real object's shapes captured by commercial 3D digitizers rather than from scratch. Then it allows the users to apply various complex 3D effects on the captured models by executing simple 2D image processing commands. Our approach solves both expressivity and operability in the 3D geometric modeling that have not been overcome in a single solution. We elaborate the internal implementation with some example 3D models created by the proposed method.