By focusing on the interior of a speaker's larynx and using infrared, he was able to convert the visible vibrations of the vocal cords into sound of fair quality, but that did not satisfy him. He worked for a while on vibrations picked up from panes of glass in windows and on framed pictures, and he experimented briefly with the diaphragms in speaker systems, intercoms and telephones. He kept on into October without stopping, and finally achieved a device that would give tinny but recognizable sound from any vibrating surface - a wall, a floor, even the speaker's own cheek or forehead.