Sythesizing high-resolution audio...

I don't remember enough information theory to know whether this is even possible, but I was wondering whether one could create a high(er)-resolution audio recording from multiple low resolution sources. Specifically, imagine a bunch of fans recording a concert using the video cameras on their cell phones. The resulting audio would be pretty lousy on a per-recording basis, but if you put them all together, would it be possible to subtract out the crowd noise and reproduce a higher fidelity version? Theoretically, it seems that the same signal processing techniques (if they exist) could be used to take multiple low resolution images and synthesize a higher resolution image.