Generation of auditory steady-state responses: linearity assessment.