That is using an array to store tones that are calculated only once so it is faster, and also I can use custom waveforms that are much more complex than the simple Sine, Square, Triangle, Sawtooth waveforms.
Okay.. That makes sense.

Edit: You might want to pre-calculate this data instead of copying it from anothe resource. Then you will be able to generate a bigger array (more octaves) and you will have more precise data. Check out the "Tone frequencie" section on this wiki-page:

http://en.wikipedia.org/wiki/Note

This looks very cool.. Audio has always been like a second hobby of mine, and i often tried to find audio-programming resources. Thanks for sharing these sources.