Model-based prediction of optogenetic sound encoding in the human cochlea by future optical cochlear implants