25 Apr
2007
25 Apr
'07
8:25 a.m.
On Wed, 2007-04-25 at 10:02 +0200, Fons Adriaensen wrote:
On Wed, Apr 25, 2007 at 09:15:35AM +0200, Jens M
Andreasen wrote:
First problem would be that there is no guarantee that all devices would
agree on that a positive signal indicates pressure and the speaker
should move outwards. There is a good probability though that it will be
that way.
Second problem, as you say, to control the entire frequency band:
That is not possible with random equipment, therefore I suggest to
carefully choose a bandwith limited signal. For instance; we have here,
at certain times, a female voice on the radio giving us detailed
information of wind and pressure all along the coast. It lasts for a
while ... Those devices that have a treble or tone control could be set
to a minimum to focus on frequency band around 1K.
What I think would be possible as an experiment
though (without
involving a budget the size of the Pope or TU-Berlin), is to place a
large amount of uniform speakers in a circle and then feed them with the
same monophonic signal. If everything works as expected, it should then
appear like the sound emerges from the center of the circular array of
speakers.
...
Would that work?
It probably would, *if* you can get all speakers to be exactly in phase
over the entire frequency band. This will be difficult at HF, and having
part of the range not focused will destroy the effect and identify the
radios as the source of the sound.
You don't even need a full circle or sphere. I'm currently involved in
a project using an array of 228 speakers / 64 channels suspended as a 3m
diameter 'chandelier' from the ceiling. It will create sound sources moving
above and around the listener's head.
Right, I was getting in that direction as well. Still thinking about
swayed arrays of standard signals (as in CD:s or FM-stereo.) for unusual
sound experinces in otherwise bland environments.
I also came to think of closing "the gap in the middle" by having two
swayed arrays of left respectively right creating a virtual center
channel slightly in front of both.Then the focus would be on the lead
singer rather than the little mandolins left and right. Perhaps a
variation on a Phillips array(sp?) could be useful here:
L-R, L, L+R, R, R-L
/jens
FA
Follie! Follie! Delirio vano รจ questo !
_______________________________________________
Linux-audio-dev mailing list
Linux-audio-dev(a)lists.linuxaudio.org
http://lists.linuxaudio.org/mailman/listinfo.cgi/linux-audio-dev
--