An unmentioned physiological consideration, remembered from a lecture long ago...
The rapid onset of the sound is an important factor. The ear has muscular reflexes that lessen the damage of loud sounds, but only if the onset of the sound takes longer than the time it takes the reflex to kick in. The toilet tank sound probably had onset approaching that of a square wave, so this neatly evaded the reflex.
Evolutionarily, one can understand why there was no need to protect the ear from loud rapid-onset sounds... those sounds don't exist widely in nature, e.g. a roaring waterfall roars continuously. Someone could smash two rocks together right next to your ear, but the fact that the rocks didn't smash your cranium is perhaps more important. :-)
To that end, Mercedes-Benz is playing around with technology to create a loud noise before the airbags deploy in a crash. The ideas is that by triggering this loud noise reflex, the ear will protect itself from the louder noises of airbag deployments.
> If an impending collision is detected, the vehicle’s sound system plays a short interference signal, which can trigger a reflex, whereby the stapedius muscle in the ears contracts, which alters the link between the eardrum and the inner ear for a split second and better protects the inner ear against high acoustic pressures, which can result from a loud crash.
The rapid onset of the sound is an important factor. The ear has muscular reflexes that lessen the damage of loud sounds, but only if the onset of the sound takes longer than the time it takes the reflex to kick in. The toilet tank sound probably had onset approaching that of a square wave, so this neatly evaded the reflex.
Evolutionarily, one can understand why there was no need to protect the ear from loud rapid-onset sounds... those sounds don't exist widely in nature, e.g. a roaring waterfall roars continuously. Someone could smash two rocks together right next to your ear, but the fact that the rocks didn't smash your cranium is perhaps more important. :-)