This book addresses the yet unanswered question of the retina’s net functionality. Focussing on the spatiotemporal aspects of retinal processing, existing retina models are analyzed with respect to their capability to i) reproduce experimental data known from neurobiology as well as ii) provide specific desired functionality. It turns out that a number of existing models can be put on a common basis: a canonical model that subsumes these models. A detailed analysis of this canonical model along with a functionally reduced derivative of it shows that both models can indeed reproduce a number of experimental data, realize some proposed functionalities, and may even in principle be realized in analog VLSI hardware, thus providing many new insights for understanding retinal processing. Apart from the exposition of the analysis, this book provides a detailed and comprehensive introduction to the neurobiology of the retina, proposed retina models, and functionalities.