Detecting the location or direction of sound with MAX/MSP?
Hi all!
I'm a student studying architecture in graduate school, and doing some acoustic research on broadway.
Since I'm new to MAX, I really don't know what MAX/MSP is capable of. (I am learning though.)
so, here's my question. I have omnidirectional mic recorder. and it collects stereo sound.
I was wondering if I can use my recording on MAX to detect the direction of sound, because that's what human ears do with two ears, and hopefully make some visualized graphic with jitter?
I would appreciate your help!
Won't work with your setup. You can't equate microphones with the human ear. The human ear, as well as being like a microphone (the eardrum part), is also a spectral encoding device (head-geometry, pinnae, cochlea and other bits), and that's where a lot of the spatial information is contained. It's also connected to a thing called a brain, which does the spatial decoding.
Aside from getting directional information, in a real-world environment you'd need to conduct sound-source separation, which has, for the past 30 years, been an active area of research with some interesting results recently coming out of AI developments.
The only realistic possibility in the max environment is if you wrote an external based on recent AI research. This would be a good 2-year masters program project, if not a PhD in itself.
i would say it depends on the sound event and also how far the microphone are placed from each other.
normally one should be able to tell the position of a sound event by measuring the delay between the channels and compare it against the loadness difference between the channels.
but it is probably only realistic with sound events such a snare drum or click sound. and the room should not be too small or very good silenced.