No User logged in, System Language: 
the mask effect
The mask effect is one of the most interesting behaviours of our ears, and it has also led to important technological improvements in music.

When we hear a loud sound, two muscles in our ears (the tensor tympani and the stapedius) contract as a reflex to the acoustic stimulus, so that the tensor tympani pulls away the hammer from the eardrum and the stapedius pulls the stapes away from the oval window. This is a defence mechanism to prevent damage caused by eventual further loud sounds. However fast these movements are, though, they take a little time to readjust to normal position. This prevents us from hearing a sound immediately after an abrupt acoustic stimulus. The curious thing is that masking takes place backwards and forwards: a loud sound can also mask a softer sound that immediately precedes it. Normally, but it depends on the kind of sounds and on how similar they are, the minimum time distance that makes us perceive two sounds distinctively is in the range of 10ms: any shorter than that and we will hear just one sound.

Let's take a look at his picture: it is the waveform of two transients. One masks the other in the first shot, then they become more distant in time and perceived as two separated sounds.


The first sound sample sounds like a single sound, but in facts is two separate transients, one masked by the other one.

Masking occurs also in frequency: two sounds very similar in their frequency content, one of which is consistently lower in amplitude, will be perceived as a single sound.

Listen to this sound: it is two sine waves, one frequency is 1KHz and the other one is 10Hz higher. At first, we can only hear the louder 1KHz. The second frequency is progressively increased but its amplitude is never raised. When the difference between the two frequencies increases, we are able to listen to the two of them as two separate sounds.

It was this phenomenon that raised the perceptive encoding technology. The widely used mp3 algorithm, reduces the size of a sound file by getting rid of what we could not hear because of the masking effect.