We determine age by voice. And then we model the sound at different periods of life. You can hear how you or your parents sounded at 5 years old. Or, for example, how your baby's voice will sound at 45.