Audiovisual texts rely on their polysemiotic nature to create audiovisual narrative. For audiences who do not have access to any one of the semiotic codes, the very essence of the narrative is compromised. The nature of these texts has changed to such an extent that they have to be re-narrativised. Within that part of the field of audiovisual translation (AVT) that aims at providing access to audiovisual texts to viewers excluded from the visual codes, audio narration (AN) is discussed as a mode that seeks to provide access through an integrated, independent narrative. This mode is suggested as an alternative to the established mode of audio description (AD), both modes being found on a descriptive-narrative continuum. The article begins by investigating the problems posed to AN by the iconicity of narrative film. It is then shown how focalisation in film manifests in a number of filmic markers that have to be substituted by linguistic markers derived from written narrative in an audio narration that is integrated with the remaining iconic codes of the soundtrack. Finally, the argument is illustrated by means of a discussion of the opening sequences of Everything is illuminated (Liev Schreiber, 2005).