Few words about the link between physics and music

Posted by: Arye_Gur on 07 January 2003

I listened to a lecture about the subject in the Internet. A lecture that was given by a professor of physics Abraham Kazir (the son of Ephraim Kazir who was a president of Israel about 20 years ago) from the University of Tel Aviv. Te University starts a project in which you can see lectures on real time by the Internet.

There are going to be several lectures about the subject, and on Monday I saw the first one. It was a great surprise because I understood that there are qualities in our hearing system that science can’t explain.
I phoned to professor Kazir to ask him few question –
He told me that the link between music and physics goes through physics and psychophysics.
We know nothing about psychophysics and we don’t know everything about the ear, which is a very complicated “device".
He gave me two examples
One for a thing we can’t measure -
A conductor can define when one violin goes wrong although he is listening at the same time to 100 players – this is a thing that Physics can’t measure or explain.

The second example for a phenomenon we can measure but can’t explain, when you hear a scream from a great distance, we know that it is a scream. When we measure a scream from a short distance we get high level of loudness while getting a very low one from the far away one, but we don’t know how the brain understands that the very low amplitude voice has the same meaning of the closer one.


I asked if it will be reasonable for him to hear a stereo manufacturer that claims for a better sound although he can’t point the physics reason why,
And he answered that is sounds logical to him.

Arye
Posted on: 08 January 2003 by i am simon 2
Arye

I think this is a fasinaiting subject that you have raised.

I studied psychology at University, and we spent quite some time studying hearing and Artifitial intelliegence hearing models.

The cortex within the brain is both highly organised and highly dynamic. The number of conections within the brains network is almost unimaginable, and it is amazing what a well organised nural network can do.

Our tutor developed an AI network that only had a couple of hundred nodes (conections) that learn to distinguish between the sound of different types of engines, ie a mtorbike a car or a van.

Studying the organisation in the network is easy but understanding exactly the purpose and function of each conection is very difficult. An AI network can do some things very well after not much learnig and this is dificult to explain.

Imagine how dificult it is to explain how the brain process music!

I look forward to reading more about these lectures.

Simon
Posted on: 08 January 2003 by Arye_Gur
Simon,

I can bring details from the next lectures.

We have many argues in the Israeli forum about the question if audiophiles are not dreamers when they describe differences in quality of sound out of immeasurable parts (like changing cables, for example).
The engineers keep saying that if you hear a difference it must be measurable.
It was amazing to me to find out that such a high authority person in physics like professor kazir is saying that it is not true, and in fact there is case when there are differences in sound that can't be measurable.
He told me that there are many descriptions of sound that musicians use and no one can measure the differences between them.


Arye