[78-L] inter-aural level differences

[DAVID BURNHAM]

Don Cox wrote:

They can't be totally irrelevant, or pan-potted stereo would not work at
all. All the instruments would seem to be in the centre.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

There were early experiments in stereo where they gathered an audience of golden 
ears to demonstrate which sound characteristics provide the most accurate 
imaging information.  Two speakers were set up and a mono sound source, 
(probably a 78), was varied in level only between the speakers, (using a 
traditional pan pot).  The audience was able to locate the source accurately to 
20 or more positions between the speakers.  Then the source was played again but 
this time the sound level from each speaker was the same but the sound from one 
speaker was delayed relative to the other speaker.  When the time was alligned, 
the sound clearly seemed to come from a position half way between the speakers. 
 When they started delaying the sound to one speaker, (at mili-second 
increments), the image became unfocused until a difference of about 20 
mili-seconds was reached, at which point the sound clearly seemed to come from 
the advanced, (un delayed), speaker.  Consequently, the conclusion was reached 
that the ear uses inter-aural level differences to locate sound sources.  

This conclusion was wrong.

Ears are designed to deal with natural sounds, not sounds from speakers.  Never 
in nature does the same sound occur at two different times.  The sounds we hear 
consist of an infinite number of mono sound sources each of which have only one 
position in space.  When the sounds from speakers are time alligned, the 
positions of the speakers themselves provide the necessary time differences.  To 
give an example, say the moving sound source is positioned halfway between the 
centre and the left speaker.  Since the speakers are alligned, the speakers 
produce the wave at the same instance and the wave front from one speaker is 
identical to the wave front from the other speaker except lower in amplitude. 
 The speakers produce circular waves and when the two wave fronts meet, energy 
is exchanged from the strong wave to the weaker wave effectively bending the 
wave front so that it has the same shape as if it had been produced by a single 
speaker at the panned position.  The necessary inter-aural time difference is 
provided by this bent wave reaching the left ear a few mili-seconds before it 
reaches the right ear.

Any sound source which is further than about 18 inches from your ears reaches 
both ears at roughly the same level.  An experiment which I often use to 
demonstrate this is to ask a subject to stop one ear as completely as possible, 
using their finger.  Then I ask them to close their eyes and I snap my fingers 
somewhere in front of them - instructing them to point at my snapping fingers. 
 They are usually surprised to find that they have pointed precisely at the 
snapping sound, even though the level to one ear is considerably reduced.

db