His next program probably did. Live and learn.

Maybe you could write a program that tries successive seeds and compares them to the sample, and then prints which seeds were the closest to the sound.