We present a technique for efficiently generating photo-realistic pictures of halos and inserting them into existing photographs. First, we describe an algorithm for producing images of halos from physical parameters either coming from existing photographs or supplied by the user. The problem with the resulting images is that they are sampled in a non-uniform way. Then, we propose a specific algorithm for reconstructing the uniform version of these images from their non-uniform sampling. Finally, we explain the complete algorithm for effectively including computer-generated halos into real photographs, thus leading to new pictures with halos looking as if they had been part of the natural scenes captured by the camera.