Sampling theory is one of the mainstays of modern signal processing and encompasses mathematical theory from harmonic analysis, functional analysis, operator theory, approximation theory and computational mathematics. It is a field with a vast amount of applications ranging from medical imaging (Magnetic Resonance Imaging and X-ray Computed Tomography) to sound engineering and image processing.The key to a successful mathematical sampling theory for use in applications is to have a model that fits the real world scenarios. And the main focus of this proposal is to emphasize the following: Due to the physical models that are the foundation of modern science one can heuristically say (from a signal processing point of view) that the world is analog (continuous-time or infinite-dimensional) whereas computer science is discrete (finite-dimensional). The gap between how we actually model the world and how we can carry out computations on a computer is a fundamental hurdle.We will in this proposal display and suggest new techniques in sampling theory that will help bridging the gap between the true model and the model used in computations. These techniques stem from recent developments in functional analysis and will ultimately provide tools that allow for improved reconstruction techniques for use in medical imaging, sound engineering and in signal processing in general.