Fig 1. Experimental details.

Patients were shown movie clips in which a Japanese female produced a lip movement of “pa,” “ka,” “ta,” or “su” with a voice speaking the syllable of “pa,” “ka,” or “ta.” In the congruent condition, the syllable spoken by the voice was congruent with the lip movement. In the incongruent conditions, the audiovisual information was mismatched: In the low-incongruent condition the voice “pa” was presented with the lip movement “ka” and in the high-incongruent condition the voice “pa” was presented with the lip movement “su”. A fixation point was presented for between 500 and 1500 ms before the onset of the movie clip. The interval between the onset of the movie clip and the onset of the audio was 1000 ms. The total length of each movie clip was 2500 ms.