My first msg hasn't come through yet, so I am resending it.
Sorry if the first one does show up.
-----------------------------------------------------------------------
At 04:52 PM 6/21/01 -0700, Chris wrote:
The big problem is power density -- these things only
have to provide
cancellation in the confines of a headset ear cup. Cancellation in the
large would obviously require much higher power levels.
It is actually more complicated. The dynamics of the room would have to
be accounted for, which makes it many times more difficult than
in the headset case. Also, to be able to cancel the noise in several
directions you need more drivers and mikes.
I once toyed with the idea of a feedback-controlled stereo system
to completely elliminate the dynamics of your listening room and
hear those of the recording hall, for example. The idea was:
when you record, you place mikes separated 3-4" apart in one of
the audience chairs, mimicking the directionality of human ears
mounted on a cranium :-) . In fact, you install several such sets
of mikes, all with a slightly different orientation wrt sound source.
Then, at home, you put on a headset
that doesn't cover your ears, but it has a second set of mikes
like the ones used for recording, and some orientation sensing
device. The stereo/DSP's job is the following: to project
on your headset's mikes sound waves that produce identical local
pressure at your ears (feedback control ideas would be used) to
those that were received at the recording hall by the set of mikes
that most closely matches your current head position.
Voila`, Panoramic Sound.
The algorithms for doing this exist already, and, while slightly
complicated for the ocasional home stereo hobbyist, they
are within the grasp of any grad student in the control/DSP/cancellation
field. They involve the estimation of the reproduction room's
characteristics, feedback control and active cancellation ideas.
Somebody must be developing this exact idea somewhere. It is
just a logical step given the state of DSP and active cancellation
technology.
Carlos.
--------------------------------------------------------------
Carlos E. Murillo-Sanchez carlos_murillo(a)nospammers.ieee.org