wizard
Member
You are probably correct in thinking that a spark would be an almost ideal point source of sound, whereas the sound of an impact would radiate from the entire surface, depending on how mechanically damped it is. I think that the peak amplitude would still be highly localized to the point of impact, so you could still locate it accurately.
I am not advocating this method. I was just trying to give you something else to think about.
I am not sure he is able to do so easily, the sound (noise) would be generated along the arc too (proportionate to the freq which the arc is generated at).
If he wants to use mechanical waves he has to go for much more frequencies, much more than 20kHz. I am suspect to the mechanical waves for the said task though.