A way to design and generate desired wave fronts using multiple sound sources is proposed. In principle, any kind of wave front can be described as a sum of orthogonal basis functions, and the wave front of a specific shape can be generated by focusing multiple sources’ energy on a single basis function. Once the sound field is decomposed into a set of orthogonal functions, we can apply a conventional focusing algorithm to generate a desired wave front. Extending the concept of energy focusing, we can also generate a wave front packet, whose energy is concentrated on a group of orthogonal functions rather than single component. This enables us to generate a wave front within a desired bound. For this purpose, we employ acoustic contrast control, which focuses sound energy on a selected group of orthogonal functions by enhancing the energy difference between the focal region and others. Various example cases demonstrate how the proposed method can generate a group of planar and spherical wave fronts propagating to different directions. [Work supported by the NRL project of KISTEP and BK21 project initiated by Ministry of Education and Human Resources Development of Korea.]