WARNING: I’m just barely understanding what I’m talking about here just yet.:
*woah* entering a technical field to attempt to understand it can be a whirlwind experience. The simplest questions can have some very accurate but “hard to digest in one sitting” answers.
So: it appears that: the Auditory Image Model (AIM) is the current model used for human sound perception research.
[how do we decode sound? It’s complicated]
According to AIM, the brain converts auditory input into “scale-shift covariant auditory images”.
Here’s one of the things that makes this powerful: From this information, we can tell how BIG the source of the resonant chamber is.
In short, we can tell if it’s a little kid making the sound, an adult making a sound, a small speaker on the computer, a large high quality speaker, the great outdoors — EVEN THOUGH we might not see the source and EVEN IF the sounds are “the same” using most other common measurements of sound.
The person who unlocked the key in the mid 90s and is still working hard on it today is http://www.neuroscience.cam.ac.uk/directory/profile.php?rdp1 Dr Roy Patterson, or RD Patterson in papers.
The Auditory Image Model (AIM) describes how humans hear sound more accurately than the usual models we use for modeling sound.
They’ve been using this in acoustic research since the mid 1990s when Dr. Roy Patterson figured it out and nobody has successfully replaced it because it works so darned well.
I wish I could do this in Audacity but it *seems* to only be available in MATLAB and C. I don’t have MATLAB but hopefully I can find someone who compiled the C for Windows… if not, I might have to get my hands a little dirty. I want to be able to do this.