Docs are a work-in-progress. Later, these pages will be all filled with music theory.

Do you know what you call a trifold plan to cure fugue addiction? A Sinfonia! Oh wait, that would be a three part invention, not intervention... Still works to cure fugue addiction...

I used to not be able to write fugues at all so I went to doctor for a cure. He prescribed something to make all of my fugues sound like Bach's. It worked so well I had to ask my doctor what he had prescribed. He said "Oh, it's just a minor tonic,"

"i" will show myself out.

Seriously, though. Bach in a Box is a pet project that is purposely avoiding using a neural network as it is impossible to truly understand what the neurons and synapses are doing. Do they TRUELY understand the music theory or have they just found a happy little local minima to settle in? You can't tell what process the neural network is using to produce what it produces. Only that what it produces sounds somewhat like something Bach could have written. Neural networks, to my knowledge, when given the task of replicating baroque music, always seem to have the same issues: producing pieces that don't "evolve" as they progress and never having a satisfying ending.

Hence, Bach in a Box will be built the hard way: by manually programming the rules of music theory (unfortunately it is currently being done by someone who doesn't understand the rules of music theory :P)