Digital Libraries can contain content stored in a variety of media - text either handwritten or machine-readable, images, audio, video, and combinations of these. Digital Libraries research has, for the most part, been concerned with capturing and organising this material, working on issues like interoperability, security, rights management, encoding standards, and other issues. But what of content-based access techniques, when we wish to find things that we're looking for in those libraries ? So for text what techniques can we use to access materials, how can we overcome the language barrier for multilingual texts, for hand-written documents what is possible, for images like photos or paintings, for video archives both older material and movies and TV content ? This talk will present a kind of overview, a state of the art summary of what we can now do with access to DL artifacts directly based on their content alone which compliments the extensive work done by others on access based on metadata.