On 03/27/2018 01:18 PM, Fons Adriaensen wrote:
On Tue, Mar 27, 2018 at 09:08:54PM +0200, Jörn Nettingsmeier wrote:
Sweet! A resounding "me too" to Stefan's question about the matrix, since
you're one channel short :)
Looking at the geometry, I guess you sacrificed the second-order
rotationally symmetric component (FuMa R or ACN 08), which seems to be a
good choice to me.
ACN 6 (= R) actually. The only possible choice given this geometry.
The actual R output is derived from W, which is correct for (near)
horizontal sources.
Hey Fons, hmmm... it is better to inject the proper amount of W into R
than to have no R at all? I can see how everything would be (slightly?)
more correct for (near) horizontal sources given the shape of R
(horizontal plane omni), but it would also alias into R anything that
comes from non-(near)horizontal sources. Presumably that does not matter
as much?
-- Fernando
The A/B processing is a combination of IIR filters and an 8 by 8
convolution matrix. The real magic is of course how this matrix
is computed. To know more you'll have to sell your soul :-)
_______________________________________________
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe here, edit
account or options, view archives and so on.
