Photoacoustic is usually done with a chopped or pulsed light source at a few hundred Hz (ie. audio frequencies). If the molecules don't absorb at that wavelength or if the molecules are not present, the light just goes through. If light is absorbed, there will be slight heating when the beam is on resulting in a sound wave at the chopping/pulsing frequency. This is detected by the microphone. For weak absorptions, the intensity of the sound wave will be linear in the concentration.

I think this is what is meant by "humming": There is a sound that is generated when the target molecule is present.