A system is disclosed for playing prerecorded audio encoded in a fault tolerant manner as a series of invisible dots carried on a substrate together with a photographic image. The system has a detector that detects the dot form of the prerecorded audio on the substrate and outputs a first signal; a decoder interconnected to the detector that decodes the first signal to produce a output signal; and an audio emitter interconnected to the processor that receives the output signal and outputs corresponding sounds. The dots may be infrared absorbing and the encoding can include Reed-Solomon encoding of the prerecorded audio. The system can include a wand-like arm having a slot through which the photograph is inserted.