How do you capture many different voices with just four values?

How do you capture many different voices with just four values?