This chapter explores audiovisual translation (AVT) in relation to models of psycholinguistics and cognitive perception. It introduces the concept of media accessibility as embracing various types of translation—namely sign language interpreting, subtitling for the deaf and hard of hearing and audio description (AD) for people who are blind or partially sighted—that enable users with a sensory disability to access audiovisual media content. L. Fryer and J. Freeman showed that AD users were more likely to feel frightened if the AD for a horror film was delivered by a human voice than by a synthesized voice. A human voice was able to add another layer of meaning beyond the semantic, through prosody. AD users can use prosody to understand a character's emotional frame of mind, not simply from the dialogue but also from the cadences that the describer or voice talent uses to deliver the AD. Mental models are important is because the AD must be congruent with the context.