To provide a voice reproduction system and electronic camera capable of correctly reading voice information by using a camera for a reader for reading the voice information in the form of optical readable information code.

When the camera (digital camera) 10 reads an optical readable voice code 14 provided with a central maker 18 to be a mark for focussing, an AF (auto focus) control means provided in the camera 10 focuses the central marker 18 as an AF detection point. When focusing is impossible by AF control, control of the camera 10 is set in photographing inhibition not to read the voice code 14. Since the voice code 14 is not read carelessly when the AF does not operate correctly, conversion malfunction of voice data is prevented from occurring due to an AF malfunction.