Here's a video of a functional prototype.
Here are a few things learned thus far:
1) The 8 ohm speaker needs to face the reader for maximum volume and clarity.
2) The speaker housing should not interfere with turning the pages.
3) Depth of the speaker will determine the depth of the housing available for electronics.
This means that speaker selection (maximum quality and volume with minimal size) will drive the enclosure construction.
Instead of buttons to select a song, photo resistors will indicate which page is being examined.
A tilt switch array can be used to sense that the book has been picked up.
An array of mechanical tilt switches will be used to put the book into "ready" mode.
Here's an overview of the project.
We are using an Adafruit Sound Board to store and replay the audio bits. The board is very small and it can handle multiple audio files. Separate pins can be activated to play individual files (up to eleven).