Hi,
Im not too sure i understand your question but here goes:
yes, Voice recognition depends entirley on sound waves. If there weren't sound waves, how could the mic pick up our voices and transmit them to the computer?
It works like this: our voices vibrate the air varying pitches (or frequencies) depending on what we are saying. The mic converts these waves into voltage signals, which the computer can use. The computer breaks these signals int small pieces, usually about 1/100th of a second each.
It takes each of these tiny 1/100th signals and obtains the frequency content of it. It uses the frequency information form the sound and compares it to a huge database that maps different frequency patterns to words, and thus figures out what word you were saying.