hardware for 20 speaker soundinstallation

From your description I get the impression that all of the speakers
will be on a straight line, thus really 1D, but I might be misunderstanding.

I used 16 speakers on a line for an installation with Kurt Ralske two
years ago. Some photos from the process of installing can be found here:

http://www.bek.no/Members/lossius/lostblog/373

For that installation I used the ambisonic external from ParisNord. When
I first started thinking of positioning the speakers this way I imagined
the row of speaker as a kind of virtual string, with sound moving back
and forth along the line. In the end the sounding result was very
different. I suspect that with the speaker setup and spatialisation
technique used, it started behave a bit similar to wavefield synthesis,
all loudspeakers contributing to a sound wave that felt to radiate as a
more or less planar perpendicular to the wall. Even if sound levels were
not all that loud, it had a very strong physical impact. It felt like
waves of sound were passing by and through me at great speed. I used the
ambisonic technique to virtually position the source of sound some
meters behind the row of speakers, moving it back and forth along the
direction of the line, also several meters past the line segment defined
by the speakers. The impression of this was that the sound sometimes
seemed to transcend the side walls of the gallery space.

While developing I also tried to use vbap in a similar wacky way, but I
never managed to get any interesting results from it. It was difficult
to locate exactly where along the speaker row the sound was coming from,
and from a few meters distance any movement along the 8 meter line
defined by the speakers didn't feel like that much of a movement after all.

Using ambisonics the effect got a lot more pronounced, and I was able to
extend and exaggerate the length of the line that sound was moving along.

If you ant to test this for yourself I would recommend that you try it
out with both the ICST ambisonic externals and the ones from parishNord.
ICST have two big advantages over ParisNord, higher order
encoding/decoding and encoding and decoding taking place in separate
externals. On the other hand, looking at the source code for both
externals, it seems to me that the ParishNord obejcts change the balance
between omni and first order signals based on the distance of the
source. In comparisment ICST externals only adjust balance as long as
the source is positioned outside the unit circle/hemisphere. That might
have influence on how sources moving into and coming from a distance are
perceived.

Best,
Trond

Sachs & Eitan PR wrote:
> while we're on the subject::
> I am working on an installation with a 2D 16 speakers linear array,
> mounted on a wall, next to which the audience can wall around freely.
>
> I am debating whether to use multiouts~ or Vbap/Rvbap for the spatial
> dispersing of the sounds.
> How would you guys compare between the two?
> thanks,
> Zax