On 2012-03-09, j.l.ander...@phonecoop.coop wrote:
Interestingly enough... this seems to imply that for HRTF decoding,
you're best choices are either
a) build a decoder using the smallest number of speakers necessary
(between 6 and 8 for FOA)
If your source signal is the ambisonic B-format in four channels,
perhsps so. But we already said so, and we're not too sure about it
because as of now the perceptual/psychoacoustic stuff underlying 4<
channels really hasn't been sorted out. In theory it ought to be better.
But that's early theory, which has been partially cancelled by newer
empirical tests.
In this circuit we at least try and do the tests. Right now I'd say that
psychoacoustically speaking, the basic 4-speaker rig is better than any
other pantophonic rig till something like a hundred equally spaced
speakers. While adding a couple of -- again not too many -- speakers
above and below can suddenly make for a periphonic system at a very few
transmission channels extra.
Any extra channels will prove counter productive if they're not combined
with with a much more interesting psychoaccoustical analysis. Over a
wider sweet spot, over its edges, and within it as well. For once
something Gerzon never did.
b) build a decoder using your complete HRTF sample set (I.e., as many
points from your set as your decoder method easily accommodates)
Come now. Marrying the free field the Ambisonic approximation to the
near field which pointwise sampling of HRTFs does... That isn't rocket
surgery. The likes of Filippo Fazi have already tackled that one. Or if
they haven't tried their hand with it, it only takes a hint from me to
show how to make it into being.
Hint: it's all about Fourier analysis over a sphere. Hint: every
optimization method used for L^2 problems applies over over the sphere
as well. In the spherical surface harmonical basis which is the
ambisonic basis as well.
Robert Greene might once again like to intervene.
--
Sampo Syreeni, aka decoy - de...@iki.fi, http://decoy.iki.fi/front
+358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2
_______________________________________________
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound
